#!/usr/bin/env python from xml.etree.cElementTree import * from os.path import basename from functools import reduce import getopt import os import sys import errno import time import re # Jump to the bottom of this file for the main routine # Some hacks to make the API more readable, and to keep backwards compability _cname_re = re.compile('([A-Z0-9][a-z]+|[A-Z0-9]+(?![a-z])|[a-z]+)') _cname_special_cases = {'DECnet':'decnet'} _extension_special_cases = ['XPrint', 'XCMisc', 'BigRequests'] _cplusplus_annoyances = {'class' : '_class', 'new' : '_new', 'delete': '_delete'} _c_keywords = {'default' : '_default'} _hlines = [] _hlevel = 0 _clines = [] _clevel = 0 _ns = None # global variable to keep track of serializers and # switch data types due to weird dependencies finished_serializers = [] finished_sizeof = [] finished_switch = [] # keeps enum objects so that we can refer to them when generating manpages. enums = {} manpaths = False def _h(fmt, *args): ''' Writes the given line to the header file. ''' _hlines[_hlevel].append(fmt % args) def _c(fmt, *args): ''' Writes the given line to the source file. ''' _clines[_clevel].append(fmt % args) def _hc(fmt, *args): ''' Writes the given line to both the header and source files. ''' _h(fmt, *args) _c(fmt, *args) # XXX See if this level thing is really necessary. def _h_setlevel(idx): ''' Changes the array that header lines are written to. Supports writing different sections of the header file. ''' global _hlevel while len(_hlines) <= idx: _hlines.append([]) _hlevel = idx def _c_setlevel(idx): ''' Changes the array that source lines are written to. Supports writing to different sections of the source file. ''' global _clevel while len(_clines) <= idx: _clines.append([]) _clevel = idx def _n_item(str): ''' Does C-name conversion on a single string fragment. Uses a regexp with some hard-coded special cases. ''' if str in _cname_special_cases: return _cname_special_cases[str] else: split = _cname_re.finditer(str) name_parts = [match.group(0) for match in split] return '_'.join(name_parts) def _cpp(str): ''' Checks for certain C++ reserved words and fixes them. ''' if str in _cplusplus_annoyances: return _cplusplus_annoyances[str] elif str in _c_keywords: return _c_keywords[str] else: return str def _ext(str): ''' Does C-name conversion on an extension name. Has some additional special cases on top of _n_item. ''' if str in _extension_special_cases: return _n_item(str).lower() else: return str.lower() def _n(list): ''' Does C-name conversion on a tuple of strings. Different behavior depending on length of tuple, extension/not extension, etc. Basically C-name converts the individual pieces, then joins with underscores. ''' if len(list) == 1: parts = list elif len(list) == 2: parts = [list[0], _n_item(list[1])] elif _ns.is_ext: parts = [list[0], _ext(list[1])] + [_n_item(i) for i in list[2:]] else: parts = [list[0]] + [_n_item(i) for i in list[1:]] return '_'.join(parts).lower() def _t(list): ''' Does C-name conversion on a tuple of strings representing a type. Same as _n but adds a "_t" on the end. ''' if len(list) == 1: parts = list elif len(list) == 2: parts = [list[0], _n_item(list[1]), 't'] elif _ns.is_ext: parts = [list[0], _ext(list[1])] + [_n_item(i) for i in list[2:]] + ['t'] else: parts = [list[0]] + [_n_item(i) for i in list[1:]] + ['t'] return '_'.join(parts).lower() def c_open(self): ''' Exported function that handles module open. Opens the files and writes out the auto-generated comment, header file includes, etc. ''' global _ns _ns = self.namespace _ns.c_ext_global_name = _n(_ns.prefix + ('id',)) # Build the type-name collision avoidance table used by c_enum build_collision_table() _h_setlevel(0) _c_setlevel(0) _hc('/*') _hc(' * This file generated automatically from %s by c_client.py.', _ns.file) _hc(' * Edit at your peril.') _hc(' */') _hc('') _h('/**') _h(' * @defgroup XCB_%s_API XCB %s API', _ns.ext_name, _ns.ext_name) _h(' * @brief %s XCB Protocol Implementation.', _ns.ext_name) _h(' * @{') _h(' **/') _h('') _h('#ifndef __%s_H', _ns.header.upper()) _h('#define __%s_H', _ns.header.upper()) _h('') _h('#include "xcb.h"') _c('#ifdef HAVE_CONFIG_H') _c('#include "config.h"') _c('#endif') _c('#include ') _c('#include ') _c('#include ') _c('#include /* for offsetof() */') _c('#include "xcbext.h"') _c('#include "%s.h"', _ns.header) _c('') _c('#define ALIGNOF(type) offsetof(struct { char dummy; type member; }, member)') if _ns.is_ext: for (n, h) in self.direct_imports: _hc('#include "%s.h"', h) _h('') _h('#ifdef __cplusplus') _h('extern "C" {') _h('#endif') if _ns.is_ext: _h('') _h('#define XCB_%s_MAJOR_VERSION %s', _ns.ext_name.upper(), _ns.major_version) _h('#define XCB_%s_MINOR_VERSION %s', _ns.ext_name.upper(), _ns.minor_version) _h(' ') #XXX _h('extern xcb_extension_t %s;', _ns.c_ext_global_name) _c('') _c('xcb_extension_t %s = { "%s", 0 };', _ns.c_ext_global_name, _ns.ext_xname) def c_close(self): ''' Exported function that handles module close. Writes out all the stored content lines, then closes the files. ''' _h_setlevel(2) _c_setlevel(2) _hc('') _h('') _h('#ifdef __cplusplus') _h('}') _h('#endif') _h('') _h('#endif') _h('') _h('/**') _h(' * @}') _h(' */') # Write header file hfile = open('%s.h' % _ns.header, 'w') for list in _hlines: for line in list: hfile.write(line) hfile.write('\n') hfile.close() # Write source file cfile = open('%s.c' % _ns.header, 'w') for list in _clines: for line in list: cfile.write(line) cfile.write('\n') cfile.close() def build_collision_table(): global namecount namecount = {} for v in module.types.values(): name = _t(v[0]) namecount[name] = (namecount.get(name) or 0) + 1 def c_enum(self, name): ''' Exported function that handles enum declarations. ''' enums[name] = self tname = _t(name) if namecount[tname] > 1: tname = _t(name + ('enum',)) _h_setlevel(0) _h('') _h('typedef enum %s {', tname) count = len(self.values) for (enam, eval) in self.values: count = count - 1 equals = ' = ' if eval != '' else '' comma = ',' if count > 0 else '' doc = '' if hasattr(self, "doc") and self.doc and enam in self.doc.fields: doc = '\n/**< %s */\n' % self.doc.fields[enam] _h(' %s%s%s%s%s', _n(name + (enam,)).upper(), equals, eval, comma, doc) _h('} %s;', tname) def _c_type_setup(self, name, postfix): ''' Sets up all the C-related state by adding additional data fields to all Field and Type objects. Here is where we figure out most of our variable and function names. Recurses into child fields and list member types. ''' # Do all the various names in advance self.c_type = _t(name + postfix) self.c_wiretype = 'char' if self.c_type == 'void' else self.c_type self.c_iterator_type = _t(name + ('iterator',)) self.c_next_name = _n(name + ('next',)) self.c_end_name = _n(name + ('end',)) self.c_request_name = _n(name) self.c_checked_name = _n(name + ('checked',)) self.c_unchecked_name = _n(name + ('unchecked',)) self.c_reply_name = _n(name + ('reply',)) self.c_reply_type = _t(name + ('reply',)) self.c_cookie_type = _t(name + ('cookie',)) self.c_reply_fds_name = _n(name + ('reply_fds',)) self.need_aux = False self.need_serialize = False self.need_sizeof = False self.c_aux_name = _n(name + ('aux',)) self.c_aux_checked_name = _n(name + ('aux', 'checked')) self.c_aux_unchecked_name = _n(name + ('aux', 'unchecked')) self.c_serialize_name = _n(name + ('serialize',)) self.c_unserialize_name = _n(name + ('unserialize',)) self.c_unpack_name = _n(name + ('unpack',)) self.c_sizeof_name = _n(name + ('sizeof',)) # special case: structs where variable size fields are followed by fixed size fields self.var_followed_by_fixed_fields = False if self.is_switch: self.need_serialize = True self.c_container = 'struct' for bitcase in self.bitcases: bitcase.c_field_name = _cpp(bitcase.field_name) bitcase_name = bitcase.field_type if bitcase.type.has_name else name _c_type_setup(bitcase.type, bitcase_name, ()) elif self.is_container: self.c_container = 'union' if self.is_union else 'struct' prev_varsized_field = None prev_varsized_offset = 0 first_field_after_varsized = None for field in self.fields: _c_type_setup(field.type, field.field_type, ()) if field.type.is_list: _c_type_setup(field.type.member, field.field_type, ()) if (field.type.nmemb is None): self.need_sizeof = True field.c_field_type = _t(field.field_type) field.c_field_const_type = ('' if field.type.nmemb == 1 else 'const ') + field.c_field_type field.c_field_name = _cpp(field.field_name) field.c_subscript = '[%d]' % field.type.nmemb if (field.type.nmemb and field.type.nmemb > 1) else '' field.c_pointer = ' ' if field.type.nmemb == 1 else '*' # correct the c_pointer field for variable size non-list types if not field.type.fixed_size() and field.c_pointer == ' ': field.c_pointer = '*' if field.type.is_list and not field.type.member.fixed_size(): field.c_pointer = '*' if field.type.is_switch: field.c_pointer = '*' field.c_field_const_type = 'const ' + field.c_field_type self.need_aux = True elif not field.type.fixed_size() and not field.type.is_bitcase: self.need_sizeof = True field.c_iterator_type = _t(field.field_type + ('iterator',)) # xcb_fieldtype_iterator_t field.c_iterator_name = _n(name + (field.field_name, 'iterator')) # xcb_container_field_iterator field.c_accessor_name = _n(name + (field.field_name,)) # xcb_container_field field.c_length_name = _n(name + (field.field_name, 'length')) # xcb_container_field_length field.c_end_name = _n(name + (field.field_name, 'end')) # xcb_container_field_end field.prev_varsized_field = prev_varsized_field field.prev_varsized_offset = prev_varsized_offset if prev_varsized_offset == 0: first_field_after_varsized = field field.first_field_after_varsized = first_field_after_varsized if field.type.fixed_size(): prev_varsized_offset += field.type.size # special case: intermixed fixed and variable size fields if prev_varsized_field is not None and not field.type.is_pad and field.wire: if not self.is_union: self.need_serialize = True self.var_followed_by_fixed_fields = True else: self.last_varsized_field = field prev_varsized_field = field prev_varsized_offset = 0 if self.var_followed_by_fixed_fields: if field.type.fixed_size(): field.prev_varsized_field = None if self.need_serialize: # when _unserialize() is wanted, create _sizeof() as well for consistency reasons self.need_sizeof = True # as switch does never appear at toplevel, # continue here with type construction if self.is_switch: if self.c_type not in finished_switch: finished_switch.append(self.c_type) # special: switch C structs get pointer fields for variable-sized members _c_complex(self) for bitcase in self.bitcases: bitcase_name = bitcase.type.name if bitcase.type.has_name else name _c_accessors(bitcase.type, bitcase_name, bitcase_name) # no list with switch as element, so no call to # _c_iterator(field.type, field_name) necessary if not self.is_bitcase: if self.need_serialize: if self.c_serialize_name not in finished_serializers: finished_serializers.append(self.c_serialize_name) _c_serialize('serialize', self) # _unpack() and _unserialize() are only needed for special cases: # switch -> unpack # special cases -> unserialize if self.is_switch or self.var_followed_by_fixed_fields: _c_serialize('unserialize', self) if self.need_sizeof: if self.c_sizeof_name not in finished_sizeof: if not module.namespace.is_ext or self.name[:2] == module.namespace.prefix: finished_sizeof.append(self.c_sizeof_name) _c_serialize('sizeof', self) # _c_type_setup() def _c_helper_absolute_name(prefix, field=None): """ turn prefix, which is a list of tuples (name, separator, Type obj) into a string representing a valid name in C (based on the context) if field is not None, append the field name as well """ prefix_str = '' for name, sep, obj in prefix: prefix_str += name if '' == sep: sep = '->' if ((obj.is_bitcase and obj.has_name) or # named bitcase (obj.is_switch and len(obj.parents)>1)): sep = '.' prefix_str += sep if field is not None: prefix_str += _cpp(field.field_name) return prefix_str # _c_absolute_name def _c_helper_field_mapping(complex_type, prefix, flat=False): """ generate absolute names, based on prefix, for all fields starting from complex_type if flat == True, nested complex types are not taken into account """ all_fields = {} if complex_type.is_switch: for b in complex_type.bitcases: if b.type.has_name: switch_name, switch_sep, switch_type = prefix[-1] bitcase_prefix = prefix + [(b.type.name[-1], '.', b.type)] else: bitcase_prefix = prefix if (True==flat and not b.type.has_name) or False==flat: all_fields.update(_c_helper_field_mapping(b.type, bitcase_prefix, flat)) else: for f in complex_type.fields: fname = _c_helper_absolute_name(prefix, f) if f.field_name in all_fields: raise Exception("field name %s has been registered before" % f.field_name) all_fields[f.field_name] = (fname, f) if f.type.is_container and flat==False: if f.type.is_bitcase and not f.type.has_name: new_prefix = prefix elif f.type.is_switch and len(f.type.parents)>1: # nested switch gets another separator new_prefix = prefix+[(f.c_field_name, '.', f.type)] else: new_prefix = prefix+[(f.c_field_name, '->', f.type)] all_fields.update(_c_helper_field_mapping(f.type, new_prefix, flat)) return all_fields # _c_field_mapping() def _c_helper_resolve_field_names (prefix): """ get field names for all objects in the prefix array """ all_fields = {} tmp_prefix = [] # look for fields in the remaining containers for idx, p in enumerate(prefix): name, sep, obj = p if ''==sep: # sep can be preset in prefix, if not, make a sensible guess sep = '.' if (obj.is_switch or obj.is_bitcase) else '->' # exception: 'toplevel' object (switch as well!) always have sep '->' sep = '->' if idx<1 else sep if not obj.is_bitcase or (obj.is_bitcase and obj.has_name): tmp_prefix.append((name, sep, obj)) all_fields.update(_c_helper_field_mapping(obj, tmp_prefix, flat=True)) return all_fields # _c_helper_resolve_field_names def get_expr_fields(self): """ get the Fields referenced by switch or list expression """ def get_expr_field_names(expr): if expr.op is None: if expr.lenfield_name is not None: return [expr.lenfield_name] else: # constant value expr return [] else: if expr.op == '~': return get_expr_field_names(expr.rhs) elif expr.op == 'popcount': return get_expr_field_names(expr.rhs) elif expr.op == 'sumof': # sumof expr references another list, # we need that list's length field here field = None for f in expr.lenfield_parent.fields: if f.field_name == expr.lenfield_name: field = f break if field is None: raise Exception("list field '%s' referenced by sumof not found" % expr.lenfield_name) # referenced list + its length field return [expr.lenfield_name] + get_expr_field_names(field.type.expr) elif expr.op == 'enumref': return [] else: return get_expr_field_names(expr.lhs) + get_expr_field_names(expr.rhs) # get_expr_field_names() # resolve the field names with the parent structure(s) unresolved_fields_names = get_expr_field_names(self.expr) # construct prefix from self prefix = [('', '', p) for p in self.parents] if self.is_container: prefix.append(('', '', self)) all_fields = _c_helper_resolve_field_names (prefix) resolved_fields_names = list(filter(lambda x: x in all_fields.keys(), unresolved_fields_names)) if len(unresolved_fields_names) != len(resolved_fields_names): raise Exception("could not resolve all fields for %s" % self.name) resolved_fields = [all_fields[n][1] for n in resolved_fields_names] return resolved_fields # get_expr_fields() def resolve_expr_fields(complex_obj): """ find expr fields appearing in complex_obj and descendents that cannot be resolved within complex_obj these are normally fields that need to be given as function parameters """ all_fields = [] expr_fields = [] unresolved = [] for field in complex_obj.fields: all_fields.append(field) if field.type.is_switch or field.type.is_list: expr_fields += get_expr_fields(field.type) if field.type.is_container: expr_fields += resolve_expr_fields(field.type) # try to resolve expr fields for e in expr_fields: if e not in all_fields and e not in unresolved: unresolved.append(e) return unresolved # resolve_expr_fields() def get_serialize_params(context, self, buffer_var='_buffer', aux_var='_aux'): """ functions like _serialize(), _unserialize(), and _unpack() sometimes need additional parameters: E.g. in order to unpack switch, extra parameters might be needed to evaluate the switch expression. This function tries to resolve all fields within a structure, and returns the unresolved fields as the list of external parameters. """ def add_param(params, param): if param not in params: params.append(param) # collect all fields into param_fields param_fields = [] wire_fields = [] for field in self.fields: if field.visible: # the field should appear as a parameter in the function call param_fields.append(field) if field.wire and not field.auto: if field.type.fixed_size() and not self.is_switch: # field in the xcb_out structure wire_fields.append(field) # fields like 'pad0' are skipped! # in case of switch, parameters always contain any fields referenced in the switch expr # we do not need any variable size fields here, as the switch data type contains both # fixed and variable size fields if self.is_switch: param_fields = get_expr_fields(self) # _serialize()/_unserialize()/_unpack() function parameters # note: don't use set() for params, it is unsorted params = [] # 1. the parameter for the void * buffer if 'serialize' == context: params.append(('void', '**', buffer_var)) elif context in ('unserialize', 'unpack', 'sizeof'): params.append(('const void', '*', buffer_var)) # 2. any expr fields that cannot be resolved within self and descendants unresolved_fields = resolve_expr_fields(self) for f in unresolved_fields: add_param(params, (f.c_field_type, '', f.c_field_name)) # 3. param_fields contain the fields necessary to evaluate the switch expr or any other fields # that do not appear in the data type struct for p in param_fields: if self.is_switch: typespec = p.c_field_const_type pointerspec = p.c_pointer add_param(params, (typespec, pointerspec, p.c_field_name)) else: if p.visible and not p.wire and not p.auto: typespec = p.c_field_type pointerspec = '' add_param(params, (typespec, pointerspec, p.c_field_name)) # 4. aux argument if 'serialize' == context: add_param(params, ('const %s' % self.c_type, '*', aux_var)) elif 'unserialize' == context: add_param(params, ('%s' % self.c_type, '**', aux_var)) elif 'unpack' == context: add_param(params, ('%s' % self.c_type, '*', aux_var)) # 5. switch contains all variable size fields as struct members # for other data types though, these have to be supplied separately # this is important for the special case of intermixed fixed and # variable size fields if not self.is_switch and 'serialize' == context: for p in param_fields: if not p.type.fixed_size(): add_param(params, (p.c_field_const_type, '*', p.c_field_name)) return (param_fields, wire_fields, params) # get_serialize_params() def _c_serialize_helper_insert_padding(context, code_lines, space, postpone): code_lines.append('%s /* insert padding */' % space) code_lines.append('%s xcb_pad = -xcb_block_len & (xcb_align_to - 1);' % space) # code_lines.append('%s printf("automatically inserting padding: %%%%d\\n", xcb_pad);' % space) code_lines.append('%s xcb_buffer_len += xcb_block_len + xcb_pad;' % space) if not postpone: code_lines.append('%s if (0 != xcb_pad) {' % space) if 'serialize' == context: code_lines.append('%s xcb_parts[xcb_parts_idx].iov_base = xcb_pad0;' % space) code_lines.append('%s xcb_parts[xcb_parts_idx].iov_len = xcb_pad;' % space) code_lines.append('%s xcb_parts_idx++;' % space) elif context in ('unserialize', 'unpack', 'sizeof'): code_lines.append('%s xcb_tmp += xcb_pad;' % space) code_lines.append('%s xcb_pad = 0;' % space) code_lines.append('%s }' % space) code_lines.append('%s xcb_block_len = 0;' % space) # keep tracking of xcb_parts entries for serialize return 1 # _c_serialize_helper_insert_padding() def _c_serialize_helper_switch(context, self, complex_name, code_lines, temp_vars, space, prefix): count = 0 switch_expr = _c_accessor_get_expr(self.expr, None) for b in self.bitcases: len_expr = len(b.type.expr) for n, expr in enumerate(b.type.expr): bitcase_expr = _c_accessor_get_expr(expr, None) # only one in the if len_expr == 1: code_lines.append(' if(%s & %s) {' % (switch_expr, bitcase_expr)) # multiple in the elif n == 0: # first code_lines.append(' if((%s & %s) ||' % (switch_expr, bitcase_expr)) elif len_expr == (n + 1): # last code_lines.append(' (%s & %s)) {' % (switch_expr, bitcase_expr)) else: # between first and last code_lines.append(' (%s & %s) ||' % (switch_expr, bitcase_expr)) b_prefix = prefix if b.type.has_name: b_prefix = prefix + [(b.c_field_name, '.', b.type)] count += _c_serialize_helper_fields(context, b.type, code_lines, temp_vars, "%s " % space, b_prefix, is_bitcase = True) code_lines.append(' }') # if 'serialize' == context: # count += _c_serialize_helper_insert_padding(context, code_lines, space, False) # elif context in ('unserialize', 'unpack', 'sizeof'): # # padding # code_lines.append('%s xcb_pad = -xcb_block_len & 3;' % space) # code_lines.append('%s xcb_buffer_len += xcb_block_len + xcb_pad;' % space) return count # _c_serialize_helper_switch def _c_serialize_helper_switch_field(context, self, field, c_switch_variable, prefix): """ handle switch by calling _serialize() or _unpack(), depending on context """ # switch is handled by this function as a special case param_fields, wire_fields, params = get_serialize_params(context, self) field_mapping = _c_helper_field_mapping(self, prefix) prefix_str = _c_helper_absolute_name(prefix) # find the parameters that need to be passed to _serialize()/_unpack(): # all switch expr fields must be given as parameters args = get_expr_fields(field.type) # length fields for variable size types in switch, normally only some of need # need to be passed as parameters switch_len_fields = resolve_expr_fields(field.type) # a switch field at this point _must_ be a bitcase field # we require that bitcases are "self-contiguous" bitcase_unresolved = resolve_expr_fields(self) if len(bitcase_unresolved) != 0: raise Exception('unresolved fields within bitcase is not supported at this point') # get the C names for the parameters c_field_names = '' for a in switch_len_fields: c_field_names += "%s, " % field_mapping[a.c_field_name][0] for a in args: c_field_names += "%s, " % field_mapping[a.c_field_name][0] # call _serialize()/_unpack() to determine the actual size if 'serialize' == context: length = "%s(&%s, %s&%s%s)" % (field.type.c_serialize_name, c_switch_variable, c_field_names, prefix_str, field.c_field_name) elif context in ('unserialize', 'unpack'): length = "%s(xcb_tmp, %s&%s%s)" % (field.type.c_unpack_name, c_field_names, prefix_str, field.c_field_name) return length # _c_serialize_helper_switch_field() def _c_serialize_helper_list_field(context, self, field, code_lines, temp_vars, space, prefix): """ helper function to cope with lists of variable length """ expr = field.type.expr prefix_str = _c_helper_absolute_name(prefix) param_fields, wire_fields, params = get_serialize_params('sizeof', self) param_names = [p[2] for p in params] expr_fields_names = [f.field_name for f in get_expr_fields(field.type)] resolved = list(filter(lambda x: x in param_names, expr_fields_names)) unresolved = list(filter(lambda x: x not in param_names, expr_fields_names)) field_mapping = {} for r in resolved: field_mapping[r] = (r, None) if len(unresolved)>0: tmp_prefix = prefix if len(tmp_prefix)==0: raise Exception("found an empty prefix while resolving expr field names for list %s", field.c_field_name) field_mapping.update(_c_helper_resolve_field_names(prefix)) resolved += list(filter(lambda x: x in field_mapping, unresolved)) unresolved = list(filter(lambda x: x not in field_mapping, unresolved)) if len(unresolved)>0: raise Exception('could not resolve the length fields required for list %s' % field.c_field_name) list_length = _c_accessor_get_expr(expr, field_mapping) # default: list with fixed size elements length = '%s * sizeof(%s)' % (list_length, field.type.member.c_wiretype) # list with variable-sized elements if not field.type.member.fixed_size(): length = '' if context in ('unserialize', 'sizeof', 'unpack'): int_i = ' unsigned int i;' xcb_tmp_len = ' unsigned int xcb_tmp_len;' if int_i not in temp_vars: temp_vars.append(int_i) if xcb_tmp_len not in temp_vars: temp_vars.append(xcb_tmp_len) # loop over all list elements and call sizeof repeatedly # this should be a bit faster than using the iterators code_lines.append("%s for(i=0; i<%s; i++) {" % (space, list_length)) code_lines.append("%s xcb_tmp_len = %s(xcb_tmp);" % (space, field.type.c_sizeof_name)) code_lines.append("%s xcb_block_len += xcb_tmp_len;" % space) code_lines.append("%s xcb_tmp += xcb_tmp_len;" % space) code_lines.append("%s }" % space) elif 'serialize' == context: code_lines.append('%s xcb_parts[xcb_parts_idx].iov_len = 0;' % space) code_lines.append('%s xcb_tmp = (char *) %s%s;' % (space, prefix_str, field.c_field_name)) code_lines.append('%s for(i=0; i<%s; i++) { ' % (space, list_length)) code_lines.append('%s xcb_block_len = %s(xcb_tmp);' % (space, field.type.c_sizeof_name)) code_lines.append('%s xcb_parts[xcb_parts_idx].iov_len += xcb_block_len;' % space) code_lines.append('%s }' % space) code_lines.append('%s xcb_block_len = xcb_parts[xcb_parts_idx].iov_len;' % space) return length # _c_serialize_helper_list_field() def _c_serialize_helper_fields_fixed_size(context, self, field, code_lines, temp_vars, space, prefix): # keep the C code a bit more readable by giving the field name if not self.is_bitcase: code_lines.append('%s /* %s.%s */' % (space, self.c_type, field.c_field_name)) else: scoped_name = [p[2].c_type if idx==0 else p[0] for idx, p in enumerate(prefix)] typename = reduce(lambda x,y: "%s.%s" % (x, y), scoped_name) code_lines.append('%s /* %s.%s */' % (space, typename, field.c_field_name)) abs_field_name = _c_helper_absolute_name(prefix, field) # default for simple cases: call sizeof() length = "sizeof(%s)" % field.c_field_type if context in ('unserialize', 'unpack', 'sizeof'): # default: simple cast value = ' %s = *(%s *)xcb_tmp;' % (abs_field_name, field.c_field_type) # padding - we could probably just ignore it if field.type.is_pad and field.type.nmemb > 1: value = '' for i in range(field.type.nmemb): code_lines.append('%s %s[%d] = *(%s *)xcb_tmp;' % (space, abs_field_name, i, field.c_field_type)) # total padding = sizeof(pad0) * nmemb length += " * %d" % field.type.nmemb if field.type.is_list: # no such case in the protocol, cannot be tested and therefore ignored for now raise Exception('list with fixed number of elemens unhandled in _unserialize()') elif 'serialize' == context: value = ' xcb_parts[xcb_parts_idx].iov_base = (char *) ' if field.type.is_expr: # need to register a temporary variable for the expression in case we know its type if field.type.c_type is None: raise Exception("type for field '%s' (expression '%s') unkown" % (field.field_name, _c_accessor_get_expr(field.type.expr))) temp_vars.append(' %s xcb_expr_%s = %s;' % (field.type.c_type, _cpp(field.field_name), _c_accessor_get_expr(field.type.expr, prefix))) value += "&xcb_expr_%s;" % _cpp(field.field_name) elif field.type.is_pad: if field.type.nmemb == 1: value += "&xcb_pad;" else: # we could also set it to 0, see definition of xcb_send_request() value = ' xcb_parts[xcb_parts_idx].iov_base = xcb_pad0;' length += "*%d" % field.type.nmemb else: # non-list type with fixed size if field.type.nmemb == 1: value += "&%s;" % (abs_field_name) # list with nmemb (fixed size) elements else: value += '%s;' % (abs_field_name) length = '%d' % field.type.nmemb return (value, length) # _c_serialize_helper_fields_fixed_size() def _c_serialize_helper_fields_variable_size(context, self, field, code_lines, temp_vars, space, prefix): prefix_str = _c_helper_absolute_name(prefix) if context in ('unserialize', 'unpack', 'sizeof'): value = '' var_field_name = 'xcb_tmp' # special case: intermixed fixed and variable size fields if self.var_followed_by_fixed_fields and 'unserialize' == context: value = ' %s = (%s *)xcb_tmp;' % (field.c_field_name, field.c_field_type) temp_vars.append(' %s *%s;' % (field.type.c_type, field.c_field_name)) # special case: switch if 'unpack' == context: value = ' %s%s = (%s *)xcb_tmp;' % (prefix_str, field.c_field_name, field.c_field_type) elif 'serialize' == context: # variable size fields appear as parameters to _serialize() if the # 'toplevel' container is not a switch prefix_string = prefix_str if prefix[0][2].is_switch else '' var_field_name = "%s%s" % (prefix_string, field.c_field_name) value = ' xcb_parts[xcb_parts_idx].iov_base = (char *) %s;' % var_field_name length = '' code_lines.append('%s /* %s */' % (space, field.c_field_name)) if field.type.is_list: if value != '': # in any context, list is already a pointer, so the default assignment is ok code_lines.append("%s%s" % (space, value)) value = '' length = _c_serialize_helper_list_field(context, self, field, code_lines, temp_vars, space, prefix) elif field.type.is_switch: value = '' if context == 'serialize': # the _serialize() function allocates the correct amount memory if given a NULL pointer value = ' xcb_parts[xcb_parts_idx].iov_base = (char *)0;' length = _c_serialize_helper_switch_field(context, self, field, 'xcb_parts[xcb_parts_idx].iov_base', prefix) else: # in all remaining special cases - call _sizeof() length = "%s(%s)" % (field.type.c_sizeof_name, var_field_name) return (value, length) # _c_serialize_helper_fields_variable_size def _c_serialize_helper_fields(context, self, code_lines, temp_vars, space, prefix, is_bitcase): count = 0 need_padding = False prev_field_was_variable = False for field in self.fields: if not field.visible: if not ((field.wire and not field.auto) or 'unserialize' == context): continue # switch/bitcase: fixed size fields must be considered explicitly if field.type.fixed_size(): if self.is_bitcase or self.var_followed_by_fixed_fields: if prev_field_was_variable and need_padding: # insert padding # count += _c_serialize_helper_insert_padding(context, code_lines, space, # self.var_followed_by_fixed_fields) prev_field_was_variable = False # prefix for fixed size fields fixed_prefix = prefix value, length = _c_serialize_helper_fields_fixed_size(context, self, field, code_lines, temp_vars, space, fixed_prefix) else: continue # fields with variable size else: if field.type.is_pad: # Variable length pad is code_lines.append('%s xcb_align_to = %d;' % (space, field.type.align)) count += _c_serialize_helper_insert_padding(context, code_lines, space, self.var_followed_by_fixed_fields) continue else: # switch/bitcase: always calculate padding before and after variable sized fields if need_padding or is_bitcase: count += _c_serialize_helper_insert_padding(context, code_lines, space, self.var_followed_by_fixed_fields) value, length = _c_serialize_helper_fields_variable_size(context, self, field, code_lines, temp_vars, space, prefix) prev_field_was_variable = True # save (un)serialization C code if '' != value: code_lines.append('%s%s' % (space, value)) if field.type.fixed_size(): if is_bitcase or self.var_followed_by_fixed_fields: # keep track of (un)serialized object's size code_lines.append('%s xcb_block_len += %s;' % (space, length)) if context in ('unserialize', 'unpack', 'sizeof'): code_lines.append('%s xcb_tmp += %s;' % (space, length)) else: # variable size objects or bitcase: # value & length might have been inserted earlier for special cases if '' != length: # special case: intermixed fixed and variable size fields if (not field.type.fixed_size() and self.var_followed_by_fixed_fields and 'unserialize' == context): temp_vars.append(' int %s_len;' % field.c_field_name) code_lines.append('%s %s_len = %s;' % (space, field.c_field_name, length)) code_lines.append('%s xcb_block_len += %s_len;' % (space, field.c_field_name)) code_lines.append('%s xcb_tmp += %s_len;' % (space, field.c_field_name)) else: code_lines.append('%s xcb_block_len += %s;' % (space, length)) # increase pointer into the byte stream accordingly if context in ('unserialize', 'sizeof', 'unpack'): code_lines.append('%s xcb_tmp += xcb_block_len;' % space) if 'serialize' == context: if '' != length: code_lines.append('%s xcb_parts[xcb_parts_idx].iov_len = %s;' % (space, length)) code_lines.append('%s xcb_parts_idx++;' % space) count += 1 code_lines.append('%s xcb_align_to = ALIGNOF(%s);' % (space, 'char' if field.c_field_type == 'void' else field.c_field_type)) need_padding = True if self.var_followed_by_fixed_fields: need_padding = False return count # _c_serialize_helper_fields() def _c_serialize_helper(context, complex_type, code_lines, temp_vars, space='', prefix=[]): # count tracks the number of fields to serialize count = 0 if hasattr(complex_type, 'type'): self = complex_type.type complex_name = complex_type.name else: self = complex_type if self.var_followed_by_fixed_fields and 'unserialize' == context: complex_name = 'xcb_out' else: complex_name = '_aux' # special case: switch is serialized by evaluating each bitcase separately if self.is_switch: count += _c_serialize_helper_switch(context, self, complex_name, code_lines, temp_vars, space, prefix) # all other data types can be evaluated one field a time else: # unserialize & fixed size fields: simply cast the buffer to the respective xcb_out type if context in ('unserialize', 'unpack', 'sizeof') and not self.var_followed_by_fixed_fields: code_lines.append('%s xcb_block_len += sizeof(%s);' % (space, self.c_type)) code_lines.append('%s xcb_tmp += xcb_block_len;' % space) code_lines.append('%s xcb_buffer_len += xcb_block_len;' % space) code_lines.append('%s xcb_block_len = 0;' % space) count += _c_serialize_helper_fields(context, self, code_lines, temp_vars, space, prefix, False) # "final padding" count += _c_serialize_helper_insert_padding(context, code_lines, space, False) return count # _c_serialize_helper() def _c_serialize(context, self): """ depending on the context variable, generate _serialize(), _unserialize(), _unpack(), or _sizeof() for the ComplexType variable self """ _h_setlevel(1) _c_setlevel(1) _hc('') # _serialize() returns the buffer size _hc('int') if self.is_switch and 'unserialize' == context: context = 'unpack' cases = { 'serialize' : self.c_serialize_name, 'unserialize' : self.c_unserialize_name, 'unpack' : self.c_unpack_name, 'sizeof' : self.c_sizeof_name } func_name = cases[context] param_fields, wire_fields, params = get_serialize_params(context, self) variable_size_fields = 0 # maximum space required for type definition of function arguments maxtypelen = 0 # determine N(variable_fields) for field in param_fields: # if self.is_switch, treat all fields as if they are variable sized if not field.type.fixed_size() or self.is_switch: variable_size_fields += 1 # determine maxtypelen for p in params: maxtypelen = max(maxtypelen, len(p[0]) + len(p[1])) # write to .c/.h indent = ' '*(len(func_name)+2) param_str = [] for p in params: typespec, pointerspec, field_name = p spacing = ' '*(maxtypelen-len(typespec)-len(pointerspec)) param_str.append("%s%s%s %s%s /**< */" % (indent, typespec, spacing, pointerspec, field_name)) # insert function name param_str[0] = "%s (%s" % (func_name, param_str[0].strip()) param_str = list(map(lambda x: "%s," % x, param_str)) for s in param_str[:-1]: _hc(s) _h("%s);" % param_str[-1].rstrip(',')) _c("%s)" % param_str[-1].rstrip(',')) _c('{') code_lines = [] temp_vars = [] prefix = [] if 'serialize' == context: if not self.is_switch and not self.var_followed_by_fixed_fields: _c(' %s *xcb_out = *_buffer;', self.c_type) _c(' unsigned int xcb_out_pad = -sizeof(%s) & 3;', self.c_type) _c(' unsigned int xcb_buffer_len = sizeof(%s) + xcb_out_pad;', self.c_type) _c(' unsigned int xcb_align_to = 0;') else: _c(' char *xcb_out = *_buffer;') _c(' unsigned int xcb_buffer_len = 0;') _c(' unsigned int xcb_align_to = 0;') prefix = [('_aux', '->', self)] aux_ptr = 'xcb_out' elif context in ('unserialize', 'unpack'): _c(' char *xcb_tmp = (char *)_buffer;') if not self.is_switch: if not self.var_followed_by_fixed_fields: _c(' const %s *_aux = (%s *)_buffer;', self.c_type, self.c_type) prefix = [('_aux', '->', self)] else: _c(' %s xcb_out;', self.c_type) prefix = [('xcb_out', '.', self)] else: aux_var = '_aux' # default for unpack: single pointer # note: unserialize not generated for switch if 'unserialize' == context: aux_var = '(*_aux)' # unserialize: double pointer (!) prefix = [(aux_var, '->', self)] aux_ptr = '*_aux' _c(' unsigned int xcb_buffer_len = 0;') _c(' unsigned int xcb_block_len = 0;') _c(' unsigned int xcb_pad = 0;') _c(' unsigned int xcb_align_to = 0;') elif 'sizeof' == context: param_names = [p[2] for p in params] if self.is_switch: # switch: call _unpack() _c(' %s _aux;', self.c_type) _c(' return %s(%s, &_aux);', self.c_unpack_name, reduce(lambda x,y: "%s, %s" % (x, y), param_names)) _c('}') return elif self.var_followed_by_fixed_fields: # special case: call _unserialize() _c(' return %s(%s, NULL);', self.c_unserialize_name, reduce(lambda x,y: "%s, %s" % (x, y), param_names)) _c('}') return else: _c(' char *xcb_tmp = (char *)_buffer;') prefix = [('_aux', '->', self)] count = _c_serialize_helper(context, self, code_lines, temp_vars, prefix=prefix) # update variable size fields (only important for context=='serialize' variable_size_fields = count if 'serialize' == context: temp_vars.append(' unsigned int xcb_pad = 0;') temp_vars.append(' char xcb_pad0[3] = {0, 0, 0};') temp_vars.append(' struct iovec xcb_parts[%d];' % count) temp_vars.append(' unsigned int xcb_parts_idx = 0;') temp_vars.append(' unsigned int xcb_block_len = 0;') temp_vars.append(' unsigned int i;') temp_vars.append(' char *xcb_tmp;') elif 'sizeof' == context: # neither switch nor intermixed fixed and variable size fields: # evaluate parameters directly if not (self.is_switch or self.var_followed_by_fixed_fields): # look if we have to declare an '_aux' variable at all if len(list(filter(lambda x: x.find('_aux')!=-1, code_lines)))>0: if not self.var_followed_by_fixed_fields: _c(' const %s *_aux = (%s *)_buffer;', self.c_type, self.c_type) else: _c(' %s *_aux = malloc(sizeof(%s));', self.c_type, self.c_type) _c(' unsigned int xcb_buffer_len = 0;') _c(' unsigned int xcb_block_len = 0;') _c(' unsigned int xcb_pad = 0;') _c(' unsigned int xcb_align_to = 0;') _c('') for t in temp_vars: _c(t) _c('') for l in code_lines: _c(l) # variable sized fields have been collected, now # allocate memory and copy everything into a continuous memory area # note: this is not necessary in case of unpack if context in ('serialize', 'unserialize'): # unserialize: check for sizeof-only invocation if 'unserialize' == context: _c('') _c(' if (NULL == _aux)') _c(' return xcb_buffer_len;') _c('') _c(' if (NULL == %s) {', aux_ptr) _c(' /* allocate memory */') _c(' %s = malloc(xcb_buffer_len);', aux_ptr) if 'serialize' == context: _c(' *_buffer = xcb_out;') _c(' }') _c('') # serialize: handle variable size fields in a loop if 'serialize' == context: if not self.is_switch and not self.var_followed_by_fixed_fields: if len(wire_fields)>0: _c(' *xcb_out = *_aux;') # copy variable size fields into the buffer if variable_size_fields > 0: # xcb_out padding if not self.is_switch and not self.var_followed_by_fixed_fields: _c(' xcb_tmp = (char*)++xcb_out;') _c(' xcb_tmp += xcb_out_pad;') else: _c(' xcb_tmp = xcb_out;') # variable sized fields _c(' for(i=0; idata;', self.c_type) if self.is_union: # FIXME - how to determine the size of a variable size union?? _c(' /* FIXME - determine the size of the union %s */', self.c_type) else: if self.need_sizeof: _c(' xcb_generic_iterator_t child;') _c(' child.data = (%s *)(((char *)R) + %s(R));', self.c_type, self.c_sizeof_name) _c(' i->index = (char *) child.data - (char *) i->data;') else: _c(' xcb_generic_iterator_t child = %s;', _c_iterator_get_end(self.last_varsized_field, 'R')) _c(' i->index = child.index;') _c(' --i->rem;') _c(' i->data = (%s *) child.data;', self.c_type) else: _c(' --i->rem;') _c(' ++i->data;') _c(' i->index += sizeof(%s);', self.c_type) _c('}') _h('') _h('/**') _h(' * Return the iterator pointing to the last element') _h(' * @param i An %s', self.c_iterator_type) _h(' * @return The iterator pointing to the last element') _h(' *') _h(' * Set the current element in the iterator to the last element.') _h(' * The member rem is set to 0. The member data points to the') _h(' * last element.') _h(' */') _c('') _hc('') _hc('/*****************************************************************************') _hc(' **') _hc(' ** xcb_generic_iterator_t %s', self.c_end_name) _hc(' ** ') _hc(' ** @param %s i', self.c_iterator_type) _hc(' ** @returns xcb_generic_iterator_t') _hc(' **') _hc(' *****************************************************************************/') _hc(' ') _hc('xcb_generic_iterator_t') _h('%s (%s i /**< */);', self.c_end_name, self.c_iterator_type) _c('%s (%s i /**< */)', self.c_end_name, self.c_iterator_type) _c('{') _c(' xcb_generic_iterator_t ret;') if self.fixed_size(): _c(' ret.data = i.data + i.rem;') _c(' ret.index = i.index + ((char *) ret.data - (char *) i.data);') _c(' ret.rem = 0;') else: _c(' while(i.rem > 0)') _c(' %s(&i);', self.c_next_name) _c(' ret.data = i.data;') _c(' ret.rem = i.rem;') _c(' ret.index = i.index;') _c(' return ret;') _c('}') def _c_accessor_get_length(expr, field_mapping=None): ''' Figures out what C code is needed to get a length field. The field_mapping parameter can be used to change the absolute name of a length field. For fields that follow a variable-length field, use the accessor. Otherwise, just reference the structure field directly. ''' lenfield_name = expr.lenfield_name if lenfield_name is not None: if field_mapping is not None: lenfield_name = field_mapping[lenfield_name][0] if expr.lenfield is not None and expr.lenfield.prev_varsized_field is not None: # special case: variable and fixed size fields are intermixed # if the lenfield is among the fixed size fields, there is no need # to call a special accessor function like return field_mapping(expr.lenfield_name) elif expr.lenfield_name is not None: return lenfield_name else: return str(expr.nmemb) def _c_accessor_get_expr(expr, field_mapping): ''' Figures out what C code is needed to get the length of a list field. The field_mapping parameter can be used to change the absolute name of a length field. Recurses for math operations. Returns bitcount for value-mask fields. Otherwise, uses the value of the length field. ''' lenexp = _c_accessor_get_length(expr, field_mapping) if expr.op == '~': return '(' + '~' + _c_accessor_get_expr(expr.rhs, field_mapping) + ')' elif expr.op == 'popcount': return 'xcb_popcount(' + _c_accessor_get_expr(expr.rhs, field_mapping) + ')' elif expr.op == 'enumref': enum_name = expr.lenfield_type.name constant_name = expr.lenfield_name c_name = _n(enum_name + (constant_name,)).upper() return c_name elif expr.op == 'sumof': # locate the referenced list object list_obj = expr.lenfield_type field = None for f in expr.lenfield_parent.fields: if f.field_name == expr.lenfield_name: field = f break if field is None: raise Exception("list field '%s' referenced by sumof not found" % expr.lenfield_name) list_name = field_mapping[field.c_field_name][0] c_length_func = "%s(%s)" % (field.c_length_name, list_name) # note: xcb_sumof() has only been defined for integers c_length_func = _c_accessor_get_expr(field.type.expr, field_mapping) return 'xcb_sumof(%s, %s)' % (list_name, c_length_func) elif expr.op != None: return ('(' + _c_accessor_get_expr(expr.lhs, field_mapping) + ' ' + expr.op + ' ' + _c_accessor_get_expr(expr.rhs, field_mapping) + ')') elif expr.bitfield: return 'xcb_popcount(' + lenexp + ')' else: return lenexp def type_pad_type(type): if type == 'void': return 'char' return type def _c_accessors_field(self, field): ''' Declares the accessor functions for a non-list field that follows a variable-length field. ''' c_type = self.c_type # special case: switch switch_obj = self if self.is_switch else None if self.is_bitcase: switch_obj = self.parents[-1] if switch_obj is not None: c_type = switch_obj.c_type if field.type.is_simple: _hc('') _hc('') _hc('/*****************************************************************************') _hc(' ** ') _hc(' ** %s %s', field.c_field_type, field.c_accessor_name) _hc(' ** ') _hc(' ** @param const %s *R', c_type) _hc(' ** @returns %s', field.c_field_type) _hc(' **') _hc(' *****************************************************************************/') _hc(' ') _hc('%s', field.c_field_type) _h('%s (const %s *R /**< */);', field.c_accessor_name, c_type) _c('%s (const %s *R /**< */)', field.c_accessor_name, c_type) _c('{') if field.prev_varsized_field is None: _c(' return (%s *) (R + 1);', field.c_field_type) else: _c(' xcb_generic_iterator_t prev = %s;', _c_iterator_get_end(field.prev_varsized_field, 'R')) _c(' return * (%s *) ((char *) prev.data + XCB_TYPE_PAD(%s, prev.index) + %d);', field.c_field_type, type_pad_type(field.first_field_after_varsized.type.c_type), field.prev_varsized_offset) _c('}') else: _hc('') _hc('') _hc('/*****************************************************************************') _hc(' **') _hc(' ** %s * %s', field.c_field_type, field.c_accessor_name) _hc(' ** ') _hc(' ** @param const %s *R', c_type) _hc(' ** @returns %s *', field.c_field_type) _hc(' **') _hc(' *****************************************************************************/') _hc(' ') if field.type.is_switch and switch_obj is None: return_type = 'void *' else: return_type = '%s *' % field.c_field_type _hc(return_type) _h('%s (const %s *R /**< */);', field.c_accessor_name, c_type) _c('%s (const %s *R /**< */)', field.c_accessor_name, c_type) _c('{') if field.prev_varsized_field is None: _c(' return (%s) (R + 1);', return_type) # note: the special case 'variable fields followed by fixed size fields' # is not of any consequence here, since the ordering gets # 'corrected' in the reply function else: _c(' xcb_generic_iterator_t prev = %s;', _c_iterator_get_end(field.prev_varsized_field, 'R')) _c(' return (%s) ((char *) prev.data + XCB_TYPE_PAD(%s, prev.index) + %d);', return_type, type_pad_type(field.first_field_after_varsized.type.c_type), field.prev_varsized_offset) _c('}') def _c_accessors_list(self, field): ''' Declares the accessor functions for a list field. Declares a direct-accessor function only if the list members are fixed size. Declares length and get-iterator functions always. ''' def get_align_pad(field): prev = field.prev_varsized_field prev_prev = field.prev_varsized_field.prev_varsized_field if (prev.type.is_pad and prev.type.align > 0 and prev_prev is not None): return (prev_prev, '((-prev.index) & (%d - 1))' % prev.type.align) else: return (prev, None) list = field.type c_type = self.c_type # special case: switch # in case of switch, 2 params have to be supplied to certain accessor functions: # 1. the anchestor object (request or reply) # 2. the (anchestor) switch object # the reason is that switch is either a child of a request/reply or nested in another switch, # so whenever we need to access a length field, we might need to refer to some anchestor type switch_obj = self if self.is_switch else None if self.is_bitcase: switch_obj = self.parents[-1] if switch_obj is not None: c_type = switch_obj.c_type params = [] fields = {} parents = self.parents if hasattr(self, 'parents') else [self] # 'R': parents[0] is always the 'toplevel' container type params.append(('const %s *R' % parents[0].c_type, parents[0])) fields.update(_c_helper_field_mapping(parents[0], [('R', '->', parents[0])], flat=True)) # auxiliary object for 'R' parameters R_obj = parents[0] if switch_obj is not None: # now look where the fields are defined that are needed to evaluate # the switch expr, and store the parent objects in accessor_params and # the fields in switch_fields # 'S': name for the 'toplevel' switch toplevel_switch = parents[1] params.append(('const %s *S' % toplevel_switch.c_type, toplevel_switch)) fields.update(_c_helper_field_mapping(toplevel_switch, [('S', '->', toplevel_switch)], flat=True)) # initialize prefix for everything "below" S prefix_str = '/* %s */ S' % toplevel_switch.name[-1] prefix = [(prefix_str, '->', toplevel_switch)] # look for fields in the remaining containers for p in parents[2:] + [self]: # the separator between parent and child is always '.' here, # because of nested switch statements if not p.is_bitcase or (p.is_bitcase and p.has_name): prefix.append((p.name[-1], '.', p)) fields.update(_c_helper_field_mapping(p, prefix, flat=True)) # auxiliary object for 'S' parameter S_obj = parents[1] _h_setlevel(1) _c_setlevel(1) if list.member.fixed_size(): idx = 1 if switch_obj is not None else 0 _hc('') _hc('') _hc('/*****************************************************************************') _hc(' **') _hc(' ** %s * %s', field.c_field_type, field.c_accessor_name) _hc(' ** ') _hc(' ** @param %s', params[idx][0]) _hc(' ** @returns %s *', field.c_field_type) _hc(' **') _hc(' *****************************************************************************/') _hc(' ') _hc('%s *', field.c_field_type) _h('%s (%s /**< */);', field.c_accessor_name, params[idx][0]) _c('%s (%s /**< */)', field.c_accessor_name, params[idx][0]) _c('{') if switch_obj is not None: _c(' return %s;', fields[field.c_field_name][0]) elif field.prev_varsized_field is None: _c(' return (%s *) (R + 1);', field.c_field_type) else: (prev_varsized_field, align_pad) = get_align_pad(field) if align_pad is None: align_pad = ('XCB_TYPE_PAD(%s, prev.index)' % type_pad_type(field.first_field_after_varsized.type.c_type)) _c(' xcb_generic_iterator_t prev = %s;', _c_iterator_get_end(prev_varsized_field, 'R')) _c(' return (%s *) ((char *) prev.data + %s + %d);', field.c_field_type, align_pad, field.prev_varsized_offset) _c('}') _hc('') _hc('') _hc('/*****************************************************************************') _hc(' **') _hc(' ** int %s', field.c_length_name) _hc(' ** ') _hc(' ** @param const %s *R', c_type) _hc(' ** @returns int') _hc(' **') _hc(' *****************************************************************************/') _hc(' ') _hc('int') if switch_obj is not None: _hc('%s (const %s *R /**< */,', field.c_length_name, R_obj.c_type) spacing = ' '*(len(field.c_length_name)+2) _h('%sconst %s *S /**< */);', spacing, S_obj.c_type) _c('%sconst %s *S /**< */)', spacing, S_obj.c_type) length = _c_accessor_get_expr(field.type.expr, fields) else: _h('%s (const %s *R /**< */);', field.c_length_name, c_type) _c('%s (const %s *R /**< */)', field.c_length_name, c_type) length = _c_accessor_get_expr(field.type.expr, fields) _c('{') _c(' return %s;', length) _c('}') if field.type.member.is_simple: _hc('') _hc('') _hc('/*****************************************************************************') _hc(' **') _hc(' ** xcb_generic_iterator_t %s', field.c_end_name) _hc(' ** ') _hc(' ** @param const %s *R', c_type) _hc(' ** @returns xcb_generic_iterator_t') _hc(' **') _hc(' *****************************************************************************/') _hc(' ') _hc('xcb_generic_iterator_t') if switch_obj is not None: _hc('%s (const %s *R /**< */,', field.c_end_name, R_obj.c_type) spacing = ' '*(len(field.c_end_name)+2) _h('%sconst %s *S /**< */);', spacing, S_obj.c_type) _c('%sconst %s *S /**< */)', spacing, S_obj.c_type) else: _h('%s (const %s *R /**< */);', field.c_end_name, c_type) _c('%s (const %s *R /**< */)', field.c_end_name, c_type) _c('{') _c(' xcb_generic_iterator_t i;') param = 'R' if switch_obj is None else 'S' if switch_obj is not None: _c(' i.data = %s + %s;', fields[field.c_field_name][0], _c_accessor_get_expr(field.type.expr, fields)) elif field.prev_varsized_field == None: _c(' i.data = ((%s *) (R + 1)) + (%s);', field.type.c_wiretype, _c_accessor_get_expr(field.type.expr, fields)) else: _c(' xcb_generic_iterator_t child = %s;', _c_iterator_get_end(field.prev_varsized_field, 'R')) _c(' i.data = ((%s *) child.data) + (%s);', field.type.c_wiretype, _c_accessor_get_expr(field.type.expr, fields)) _c(' i.rem = 0;') _c(' i.index = (char *) i.data - (char *) %s;', param) _c(' return i;') _c('}') else: _hc('') _hc('') _hc('/*****************************************************************************') _hc(' **') _hc(' ** %s %s', field.c_iterator_type, field.c_iterator_name) _hc(' ** ') _hc(' ** @param const %s *R', c_type) _hc(' ** @returns %s', field.c_iterator_type) _hc(' **') _hc(' *****************************************************************************/') _hc(' ') _hc('%s', field.c_iterator_type) if switch_obj is not None: _hc('%s (const %s *R /**< */,', field.c_iterator_name, R_obj.c_type) spacing = ' '*(len(field.c_iterator_name)+2) _h('%sconst %s *S /**< */);', spacing, S_obj.c_type) _c('%sconst %s *S /**< */)', spacing, S_obj.c_type) else: _h('%s (const %s *R /**< */);', field.c_iterator_name, c_type) _c('%s (const %s *R /**< */)', field.c_iterator_name, c_type) _c('{') _c(' %s i;', field.c_iterator_type) if switch_obj is not None: _c(' i.data = %s;', fields[field.c_field_name][0]) _c(' i.rem = %s;', _c_accessor_get_expr(field.type.expr, fields)) elif field.prev_varsized_field == None: _c(' i.data = (%s *) (R + 1);', field.c_field_type) else: (prev_varsized_field, align_pad) = get_align_pad(field) if align_pad is None: align_pad = ('XCB_TYPE_PAD(%s, prev.index)' % type_pad_type(field.c_field_type)) _c(' xcb_generic_iterator_t prev = %s;', _c_iterator_get_end(prev_varsized_field, 'R')) _c(' i.data = (%s *) ((char *) prev.data + %s);', field.c_field_type, align_pad) if switch_obj is None: _c(' i.rem = %s;', _c_accessor_get_expr(field.type.expr, fields)) _c(' i.index = (char *) i.data - (char *) %s;', 'R' if switch_obj is None else 'S' ) _c(' return i;') _c('}') def _c_accessors(self, name, base): ''' Declares the accessor functions for the fields of a structure. ''' # no accessors for switch itself - # switch always needs to be unpacked explicitly # if self.is_switch: # pass # else: if True: for field in self.fields: if not field.type.is_pad: if field.type.is_list and not field.type.fixed_size(): _c_accessors_list(self, field) elif field.prev_varsized_field is not None or not field.type.fixed_size(): _c_accessors_field(self, field) def c_simple(self, name): ''' Exported function that handles cardinal type declarations. These are types which are typedef'd to one of the CARDx's, char, float, etc. ''' _c_type_setup(self, name, ()) if (self.name != name): # Typedef _h_setlevel(0) my_name = _t(name) _h('') _h('typedef %s %s;', _t(self.name), my_name) # Iterator _c_iterator(self, name) def _c_complex(self, force_packed = False): ''' Helper function for handling all structure types. Called for all structs, requests, replies, events, errors. ''' _h_setlevel(0) _h('') _h('/**') _h(' * @brief %s', self.c_type) _h(' **/') _h('typedef %s %s {', self.c_container, self.c_type) struct_fields = [] maxtypelen = 0 varfield = None for field in self.fields: if not field.type.fixed_size() and not self.is_switch and not self.is_union: varfield = field.c_field_name continue if field.wire: struct_fields.append(field) for field in struct_fields: length = len(field.c_field_type) # account for '*' pointer_spec if not field.type.fixed_size() and not self.is_union: length += 1 maxtypelen = max(maxtypelen, length) def _c_complex_field(self, field, space=''): if (field.type.fixed_size() or self.is_union or # in case of switch with switch children, don't make the field a pointer # necessary for unserialize to work (self.is_switch and field.type.is_switch)): spacing = ' ' * (maxtypelen - len(field.c_field_type)) _h('%s %s%s %s%s; /**< */', space, field.c_field_type, spacing, field.c_field_name, field.c_subscript) else: spacing = ' ' * (maxtypelen - (len(field.c_field_type) + 1)) _h('%s %s%s *%s%s; /**< */', space, field.c_field_type, spacing, field.c_field_name, field.c_subscript) if not self.is_switch: for field in struct_fields: _c_complex_field(self, field) else: for b in self.bitcases: space = '' if b.type.has_name: _h(' struct _%s {', b.c_field_name) space = ' ' for field in b.type.fields: _c_complex_field(self, field, space) if b.type.has_name: _h(' } %s;', b.c_field_name) _h('} %s%s;', 'XCB_PACKED ' if force_packed else '', self.c_type) def c_struct(self, name): ''' Exported function that handles structure declarations. ''' _c_type_setup(self, name, ()) _c_complex(self) _c_accessors(self, name, name) _c_iterator(self, name) def c_union(self, name): ''' Exported function that handles union declarations. ''' _c_type_setup(self, name, ()) _c_complex(self) _c_iterator(self, name) def _c_request_helper(self, name, cookie_type, void, regular, aux=False, reply_fds=False): ''' Declares a request function. ''' # Four stunningly confusing possibilities here: # # Void Non-void # ------------------------------ # "req" "req" # 0 flag CHECKED flag Normal Mode # void_cookie req_cookie # ------------------------------ # "req_checked" "req_unchecked" # CHECKED flag 0 flag Abnormal Mode # void_cookie req_cookie # ------------------------------ # Whether we are _checked or _unchecked checked = void and not regular unchecked = not void and not regular # What kind of cookie we return func_cookie = 'xcb_void_cookie_t' if void else self.c_cookie_type # What flag is passed to xcb_request func_flags = '0' if (void and regular) or (not void and not regular) else 'XCB_REQUEST_CHECKED' if reply_fds: if func_flags == '0': func_flags = 'XCB_REQUEST_REPLY_FDS' else: func_flags = func_flags + '|XCB_REQUEST_REPLY_FDS' # Global extension id variable or NULL for xproto func_ext_global = '&' + _ns.c_ext_global_name if _ns.is_ext else '0' # What our function name is func_name = self.c_request_name if not aux else self.c_aux_name if checked: func_name = self.c_checked_name if not aux else self.c_aux_checked_name if unchecked: func_name = self.c_unchecked_name if not aux else self.c_aux_unchecked_name param_fields = [] wire_fields = [] maxtypelen = len('xcb_connection_t') serial_fields = [] # special case: list with variable size elements list_with_var_size_elems = False for field in self.fields: if field.visible: # The field should appear as a call parameter param_fields.append(field) if field.wire and not field.auto: # We need to set the field up in the structure wire_fields.append(field) if field.type.need_serialize or field.type.need_sizeof: serial_fields.append(field) for field in param_fields: c_field_const_type = field.c_field_const_type if field.type.need_serialize and not aux: c_field_const_type = "const void" if len(c_field_const_type) > maxtypelen: maxtypelen = len(c_field_const_type) if field.type.is_list and not field.type.member.fixed_size(): list_with_var_size_elems = True _h_setlevel(1) _c_setlevel(1) _h('') _h('/**') if hasattr(self, "doc") and self.doc: if self.doc.brief: _h(' * @brief ' + self.doc.brief) else: _h(' * No brief doc yet') _h(' *') _h(' * @param c The connection') param_names = [f.c_field_name for f in param_fields] if hasattr(self, "doc") and self.doc: for field in param_fields: # XXX: hard-coded until we fix xproto.xml base_func_name = self.c_request_name if not aux else self.c_aux_name if base_func_name == 'xcb_change_gc' and field.c_field_name == 'value_mask': field.enum = 'GC' elif base_func_name == 'xcb_change_window_attributes' and field.c_field_name == 'value_mask': field.enum = 'CW' elif base_func_name == 'xcb_create_window' and field.c_field_name == 'value_mask': field.enum = 'CW' if field.enum: # XXX: why the 'xcb' prefix? key = ('xcb', field.enum) tname = _t(key) if namecount[tname] > 1: tname = _t(key + ('enum',)) _h(' * @param %s A bitmask of #%s values.' % (field.c_field_name, tname)) if self.doc and field.field_name in self.doc.fields: desc = self.doc.fields[field.field_name] for name in param_names: desc = desc.replace('`%s`' % name, '\\a %s' % (name)) desc = desc.split("\n") desc = [line if line != '' else '\\n' for line in desc] _h(' * @param %s %s' % (field.c_field_name, "\n * ".join(desc))) # If there is no documentation yet, we simply don't generate an # @param tag. Doxygen will then warn about missing documentation. _h(' * @return A cookie') _h(' *') if hasattr(self, "doc") and self.doc: if self.doc.description: desc = self.doc.description for name in param_names: desc = desc.replace('`%s`' % name, '\\a %s' % (name)) desc = desc.split("\n") _h(' * ' + "\n * ".join(desc)) else: _h(' * No description yet') else: _h(' * Delivers a request to the X server.') _h(' * ') if checked: _h(' * This form can be used only if the request will not cause') _h(' * a reply to be generated. Any returned error will be') _h(' * saved for handling by xcb_request_check().') if unchecked: _h(' * This form can be used only if the request will cause') _h(' * a reply to be generated. Any returned error will be') _h(' * placed in the event queue.') _h(' */') _c('') _hc('') _hc('/*****************************************************************************') _hc(' **') _hc(' ** %s %s', cookie_type, func_name) _hc(' ** ') spacing = ' ' * (maxtypelen - len('xcb_connection_t')) _hc(' ** @param xcb_connection_t%s *c', spacing) for field in param_fields: c_field_const_type = field.c_field_const_type if field.type.need_serialize and not aux: c_field_const_type = "const void" spacing = ' ' * (maxtypelen - len(c_field_const_type)) _hc(' ** @param %s%s %s%s', c_field_const_type, spacing, field.c_pointer, field.c_field_name) _hc(' ** @returns %s', cookie_type) _hc(' **') _hc(' *****************************************************************************/') _hc(' ') _hc('%s', cookie_type) spacing = ' ' * (maxtypelen - len('xcb_connection_t')) comma = ',' if len(param_fields) else ');' _h('%s (xcb_connection_t%s *c /**< */%s', func_name, spacing, comma) comma = ',' if len(param_fields) else ')' _c('%s (xcb_connection_t%s *c /**< */%s', func_name, spacing, comma) func_spacing = ' ' * (len(func_name) + 2) count = len(param_fields) for field in param_fields: count = count - 1 c_field_const_type = field.c_field_const_type c_pointer = field.c_pointer if field.type.need_serialize and not aux: c_field_const_type = "const void" c_pointer = '*' spacing = ' ' * (maxtypelen - len(c_field_const_type)) comma = ',' if count else ');' _h('%s%s%s %s%s /**< */%s', func_spacing, c_field_const_type, spacing, c_pointer, field.c_field_name, comma) comma = ',' if count else ')' _c('%s%s%s %s%s /**< */%s', func_spacing, c_field_const_type, spacing, c_pointer, field.c_field_name, comma) count = 2 if not self.var_followed_by_fixed_fields: for field in param_fields: if not field.type.fixed_size(): count = count + 2 if field.type.need_serialize: # _serialize() keeps track of padding automatically count -= 1 dimension = count + 2 _c('{') _c(' static const xcb_protocol_request_t xcb_req = {') _c(' /* count */ %d,', count) _c(' /* ext */ %s,', func_ext_global) _c(' /* opcode */ %s,', self.c_request_name.upper()) _c(' /* isvoid */ %d', 1 if void else 0) _c(' };') _c(' ') _c(' struct iovec xcb_parts[%d];', dimension) _c(' %s xcb_ret;', func_cookie) _c(' %s xcb_out;', self.c_type) if self.var_followed_by_fixed_fields: _c(' /* in the protocol description, variable size fields are followed by fixed size fields */') _c(' void *xcb_aux = 0;') for idx, f in enumerate(serial_fields): if aux: _c(' void *xcb_aux%d = 0;' % (idx)) if list_with_var_size_elems: _c(' unsigned int i;') _c(' unsigned int xcb_tmp_len;') _c(' char *xcb_tmp;') _c(' ') # simple request call tracing # _c(' printf("in function %s\\n");' % func_name) # fixed size fields for field in wire_fields: if field.type.fixed_size(): if field.type.is_expr: _c(' xcb_out.%s = %s;', field.c_field_name, _c_accessor_get_expr(field.type.expr, None)) elif field.type.is_pad: if field.type.nmemb == 1: _c(' xcb_out.%s = 0;', field.c_field_name) else: _c(' memset(xcb_out.%s, 0, %d);', field.c_field_name, field.type.nmemb) else: if field.type.nmemb == 1: _c(' xcb_out.%s = %s;', field.c_field_name, field.c_field_name) else: _c(' memcpy(xcb_out.%s, %s, %d);', field.c_field_name, field.c_field_name, field.type.nmemb) def get_serialize_args(type_obj, c_field_name, aux_var, context='serialize'): serialize_args = get_serialize_params(context, type_obj, c_field_name, aux_var)[2] return reduce(lambda x,y: "%s, %s" % (x,y), [a[2] for a in serialize_args]) # calls in order to free dyn. all. memory free_calls = [] _c(' ') if not self.var_followed_by_fixed_fields: _c(' xcb_parts[2].iov_base = (char *) &xcb_out;') _c(' xcb_parts[2].iov_len = sizeof(xcb_out);') _c(' xcb_parts[3].iov_base = 0;') _c(' xcb_parts[3].iov_len = -xcb_parts[2].iov_len & 3;') count = 4 for field in param_fields: if not field.type.fixed_size(): _c(' /* %s %s */', field.type.c_type, field.c_field_name) # default: simple cast to char * if not field.type.need_serialize and not field.type.need_sizeof: _c(' xcb_parts[%d].iov_base = (char *) %s;', count, field.c_field_name) if field.type.is_list: if field.type.member.fixed_size(): _c(' xcb_parts[%d].iov_len = %s * sizeof(%s);', count, _c_accessor_get_expr(field.type.expr, None), field.type.member.c_wiretype) else: list_length = _c_accessor_get_expr(field.type.expr, None) length = '' _c(" xcb_parts[%d].iov_len = 0;" % count) _c(" xcb_tmp = (char *)%s;", field.c_field_name) _c(" for(i=0; i<%s; i++) {" % list_length) _c(" xcb_tmp_len = %s(xcb_tmp);" % (field.type.c_sizeof_name)) _c(" xcb_parts[%d].iov_len += xcb_tmp_len;" % count) _c(" xcb_tmp += xcb_tmp_len;") _c(" }") else: # not supposed to happen raise Exception("unhandled variable size field %s" % field.c_field_name) else: if not aux: _c(' xcb_parts[%d].iov_base = (char *) %s;', count, field.c_field_name) idx = serial_fields.index(field) aux_var = '&xcb_aux%d' % idx context = 'serialize' if aux else 'sizeof' _c(' xcb_parts[%d].iov_len = ', count) if aux: serialize_args = get_serialize_args(field.type, aux_var, field.c_field_name, context) _c(' %s (%s);', field.type.c_serialize_name, serialize_args) _c(' xcb_parts[%d].iov_base = xcb_aux%d;' % (count, idx)) free_calls.append(' free(xcb_aux%d);' % idx) else: serialize_args = get_serialize_args(field.type, field.c_field_name, aux_var, context) func_name = field.type.c_sizeof_name _c(' %s (%s);', func_name, serialize_args) count += 1 if not (field.type.need_serialize or field.type.need_sizeof): # the _serialize() function keeps track of padding automatically _c(' xcb_parts[%d].iov_base = 0;', count) _c(' xcb_parts[%d].iov_len = -xcb_parts[%d].iov_len & 3;', count, count-1) count += 1 # elif self.var_followed_by_fixed_fields: else: _c(' xcb_parts[2].iov_base = (char *) &xcb_out;') # request header: opcodes + length _c(' xcb_parts[2].iov_len = 2*sizeof(uint8_t) + sizeof(uint16_t);') count += 1 # call _serialize() buffer_var = '&xcb_aux' serialize_args = get_serialize_args(self, buffer_var, '&xcb_out', 'serialize') _c(' xcb_parts[%d].iov_len = %s (%s);', count, self.c_serialize_name, serialize_args) _c(' xcb_parts[%d].iov_base = (char *) xcb_aux;', count) free_calls.append(' free(xcb_aux);') # no padding necessary - _serialize() keeps track of padding automatically _c(' ') for field in param_fields: if field.isfd: _c(' xcb_send_fd(c, %s);', field.c_field_name) _c(' xcb_ret.sequence = xcb_send_request(c, %s, xcb_parts + 2, &xcb_req);', func_flags) # free dyn. all. data, if any for f in free_calls: _c(f) _c(' return xcb_ret;') _c('}') def _c_reply(self, name): ''' Declares the function that returns the reply structure. ''' spacing1 = ' ' * (len(self.c_cookie_type) - len('xcb_connection_t')) spacing2 = ' ' * (len(self.c_cookie_type) - len('xcb_generic_error_t')) spacing3 = ' ' * (len(self.c_reply_name) + 2) # check if _unserialize() has to be called for any field def look_for_special_cases(complex_obj): unserialize_fields = [] # no unserialize call in case of switch if not complex_obj.is_switch: for field in complex_obj.fields: # three cases: 1. field with special case # 2. container that contains special case field # 3. list with special case elements if field.type.var_followed_by_fixed_fields: unserialize_fields.append(field) elif field.type.is_container: unserialize_fields += look_for_special_cases(field.type) elif field.type.is_list: if field.type.member.var_followed_by_fixed_fields: unserialize_fields.append(field) if field.type.member.is_container: unserialize_fields += look_for_special_cases(field.type.member) return unserialize_fields unserialize_fields = look_for_special_cases(self.reply) _h('') _h('/**') _h(' * Return the reply') _h(' * @param c The connection') _h(' * @param cookie The cookie') _h(' * @param e The xcb_generic_error_t supplied') _h(' *') _h(' * Returns the reply of the request asked by') _h(' * ') _h(' * The parameter @p e supplied to this function must be NULL if') _h(' * %s(). is used.', self.c_unchecked_name) _h(' * Otherwise, it stores the error if any.') _h(' *') _h(' * The returned value must be freed by the caller using free().') _h(' */') _c('') _hc('') _hc('/*****************************************************************************') _hc(' **') _hc(' ** %s * %s', self.c_reply_type, self.c_reply_name) _hc(' ** ') _hc(' ** @param xcb_connection_t%s *c', spacing1) _hc(' ** @param %s cookie', self.c_cookie_type) _hc(' ** @param xcb_generic_error_t%s **e', spacing2) _hc(' ** @returns %s *', self.c_reply_type) _hc(' **') _hc(' *****************************************************************************/') _hc(' ') _hc('%s *', self.c_reply_type) _hc('%s (xcb_connection_t%s *c /**< */,', self.c_reply_name, spacing1) _hc('%s%s cookie /**< */,', spacing3, self.c_cookie_type) _h('%sxcb_generic_error_t%s **e /**< */);', spacing3, spacing2) _c('%sxcb_generic_error_t%s **e /**< */)', spacing3, spacing2) _c('{') if len(unserialize_fields)>0: # certain variable size fields need to be unserialized explicitly _c(' %s *reply = (%s *) xcb_wait_for_reply(c, cookie.sequence, e);', self.c_reply_type, self.c_reply_type) _c(' int i;') for field in unserialize_fields: if field.type.is_list: _c(' %s %s_iter = %s(reply);', field.c_iterator_type, field.c_field_name, field.c_iterator_name) _c(' int %s_len = %s(reply);', field.c_field_name, field.c_length_name) _c(' %s *%s_data;', field.c_field_type, field.c_field_name) else: raise Exception('not implemented: call _unserialize() in reply for non-list type %s', field.c_field_type) # call _unserialize(), using the reply as source and target buffer _c(' /* special cases: transform parts of the reply to match XCB data structures */') for field in unserialize_fields: if field.type.is_list: _c(' for(i=0; i<%s_len; i++) {', field.c_field_name) _c(' %s_data = %s_iter.data;', field.c_field_name, field.c_field_name) _c(' %s((const void *)%s_data, &%s_data);', field.type.c_unserialize_name, field.c_field_name, field.c_field_name) _c(' %s(&%s_iter);', field.type.c_next_name, field.c_field_name) _c(' }') # return the transformed reply _c(' return reply;') else: _c(' return (%s *) xcb_wait_for_reply(c, cookie.sequence, e);', self.c_reply_type) _c('}') def _c_reply_has_fds(self): for field in self.fields: if field.isfd: return True return False def _c_reply_fds(self, name): ''' Declares the function that returns fds related to the reply. ''' spacing1 = ' ' * (len(self.c_reply_type) - len('xcb_connection_t')) spacing3 = ' ' * (len(self.c_reply_fds_name) + 2) _h('') _h('/**') _h(' * Return the reply fds') _h(' * @param c The connection') _h(' * @param reply The reply') _h(' *') _h(' * Returns the array of reply fds of the request asked by') _h(' * ') _h(' * The returned value must be freed by the caller using free().') _h(' */') _c('') _hc('') _hc('/*****************************************************************************') _hc(' **') _hc(' ** int * %s', self.c_reply_fds_name) _hc(' ** ') _hc(' ** @param xcb_connection_t%s *c', spacing1) _hc(' ** @param %s *reply', self.c_reply_type) _hc(' ** @returns int *') _hc(' **') _hc(' *****************************************************************************/') _hc(' ') _hc('int *') _hc('%s (xcb_connection_t%s *c /**< */,', self.c_reply_fds_name, spacing1) _h('%s%s *reply /**< */);', spacing3, self.c_reply_type) _c('%s%s *reply /**< */)', spacing3, self.c_reply_type) _c('{') _c(' return xcb_get_reply_fds(c, reply, sizeof(%s) + 4 * reply->length);', self.c_reply_type) _c('}') def _c_opcode(name, opcode): ''' Declares the opcode define for requests, events, and errors. ''' _h_setlevel(0) _h('') _h('/** Opcode for %s. */', _n(name)) _h('#define %s %s', _n(name).upper(), opcode) def _c_cookie(self, name): ''' Declares the cookie type for a non-void request. ''' _h_setlevel(0) _h('') _h('/**') _h(' * @brief %s', self.c_cookie_type) _h(' **/') _h('typedef struct %s {', self.c_cookie_type) _h(' unsigned int sequence; /**< */') _h('} %s;', self.c_cookie_type) def _man_request(self, name, cookie_type, void, aux): param_fields = [f for f in self.fields if f.visible] func_name = self.c_request_name if not aux else self.c_aux_name def create_link(linkname): name = 'man/%s.%s' % (linkname, section) if manpaths: sys.stdout.write(name) f = open(name, 'w') f.write('.so man%s/%s.%s' % (section, func_name, section)) f.close() if manpaths: sys.stdout.write('man/%s.%s ' % (func_name, section)) # Our CWD is src/, so this will end up in src/man/ f = open('man/%s.%s' % (func_name, section), 'w') f.write('.TH %s %s "%s" "%s" "XCB Requests"\n' % (func_name, section, center_footer, left_footer)) # Left-adjust instead of adjusting to both sides f.write('.ad l\n') f.write('.SH NAME\n') brief = self.doc.brief if hasattr(self, "doc") and self.doc else '' f.write('%s \\- %s\n' % (func_name, brief)) f.write('.SH SYNOPSIS\n') # Don't split words (hyphenate) f.write('.hy 0\n') f.write('.B #include \n' % _ns.header) # function prototypes prototype = '' count = len(param_fields) for field in param_fields: count = count - 1 c_field_const_type = field.c_field_const_type c_pointer = field.c_pointer if c_pointer == ' ': c_pointer = '' if field.type.need_serialize and not aux: c_field_const_type = "const void" c_pointer = '*' comma = ', ' if count else ');' prototype += '%s\\ %s\\fI%s\\fP%s' % (c_field_const_type, c_pointer, field.c_field_name, comma) f.write('.SS Request function\n') f.write('.HP\n') base_func_name = self.c_request_name if not aux else self.c_aux_name f.write('%s \\fB%s\\fP(xcb_connection_t\\ *\\fIconn\\fP, %s\n' % (cookie_type, base_func_name, prototype)) create_link('%s_%s' % (base_func_name, ('checked' if void else 'unchecked'))) if not void: f.write('.PP\n') f.write('.SS Reply datastructure\n') f.write('.nf\n') f.write('.sp\n') f.write('typedef %s %s {\n' % (self.reply.c_container, self.reply.c_type)) struct_fields = [] maxtypelen = 0 for field in self.reply.fields: if not field.type.fixed_size() and not self.is_switch and not self.is_union: continue if field.wire: struct_fields.append(field) for field in struct_fields: length = len(field.c_field_type) # account for '*' pointer_spec if not field.type.fixed_size(): length += 1 maxtypelen = max(maxtypelen, length) def _c_complex_field(self, field, space=''): if (field.type.fixed_size() or # in case of switch with switch children, don't make the field a pointer # necessary for unserialize to work (self.is_switch and field.type.is_switch)): spacing = ' ' * (maxtypelen - len(field.c_field_type)) f.write('%s %s%s \\fI%s\\fP%s;\n' % (space, field.c_field_type, spacing, field.c_field_name, field.c_subscript)) else: spacing = ' ' * (maxtypelen - (len(field.c_field_type) + 1)) f.write('ELSE %s = %s\n' % (field.c_field_type, field.c_field_name)) #_h('%s %s%s *%s%s; /**< */', space, field.c_field_type, spacing, field.c_field_name, field.c_subscript) if not self.is_switch: for field in struct_fields: _c_complex_field(self, field) else: for b in self.bitcases: space = '' if b.type.has_name: space = ' ' for field in b.type.fields: _c_complex_field(self, field, space) if b.type.has_name: print >> sys.stderr, 'ERROR: New unhandled documentation case' pass f.write('} \\fB%s\\fP;\n' % self.reply.c_type) f.write('.fi\n') f.write('.SS Reply function\n') f.write('.HP\n') f.write(('%s *\\fB%s\\fP(xcb_connection_t\\ *\\fIconn\\fP, %s\\ ' '\\fIcookie\\fP, xcb_generic_error_t\\ **\\fIe\\fP);\n') % (self.c_reply_type, self.c_reply_name, self.c_cookie_type)) create_link('%s' % self.c_reply_name) has_accessors = False for field in self.reply.fields: if field.type.is_list and not field.type.fixed_size(): has_accessors = True elif field.prev_varsized_field is not None or not field.type.fixed_size(): has_accessors = True if has_accessors: f.write('.SS Reply accessors\n') def _c_accessors_field(self, field): ''' Declares the accessor functions for a non-list field that follows a variable-length field. ''' c_type = self.c_type # special case: switch switch_obj = self if self.is_switch else None if self.is_bitcase: switch_obj = self.parents[-1] if switch_obj is not None: c_type = switch_obj.c_type if field.type.is_simple: f.write('%s %s (const %s *reply)\n' % (field.c_field_type, field.c_accessor_name, c_type)) create_link('%s' % field.c_accessor_name) else: f.write('%s *%s (const %s *reply)\n' % (field.c_field_type, field.c_accessor_name, c_type)) create_link('%s' % field.c_accessor_name) def _c_accessors_list(self, field): ''' Declares the accessor functions for a list field. Declares a direct-accessor function only if the list members are fixed size. Declares length and get-iterator functions always. ''' list = field.type c_type = self.reply.c_type # special case: switch # in case of switch, 2 params have to be supplied to certain accessor functions: # 1. the anchestor object (request or reply) # 2. the (anchestor) switch object # the reason is that switch is either a child of a request/reply or nested in another switch, # so whenever we need to access a length field, we might need to refer to some anchestor type switch_obj = self if self.is_switch else None if self.is_bitcase: switch_obj = self.parents[-1] if switch_obj is not None: c_type = switch_obj.c_type params = [] fields = {} parents = self.parents if hasattr(self, 'parents') else [self] # 'R': parents[0] is always the 'toplevel' container type params.append(('const %s *\\fIreply\\fP' % parents[0].c_type, parents[0])) fields.update(_c_helper_field_mapping(parents[0], [('R', '->', parents[0])], flat=True)) # auxiliary object for 'R' parameters R_obj = parents[0] if switch_obj is not None: # now look where the fields are defined that are needed to evaluate # the switch expr, and store the parent objects in accessor_params and # the fields in switch_fields # 'S': name for the 'toplevel' switch toplevel_switch = parents[1] params.append(('const %s *S' % toplevel_switch.c_type, toplevel_switch)) fields.update(_c_helper_field_mapping(toplevel_switch, [('S', '->', toplevel_switch)], flat=True)) # initialize prefix for everything "below" S prefix_str = '/* %s */ S' % toplevel_switch.name[-1] prefix = [(prefix_str, '->', toplevel_switch)] # look for fields in the remaining containers for p in parents[2:] + [self]: # the separator between parent and child is always '.' here, # because of nested switch statements if not p.is_bitcase or (p.is_bitcase and p.has_name): prefix.append((p.name[-1], '.', p)) fields.update(_c_helper_field_mapping(p, prefix, flat=True)) # auxiliary object for 'S' parameter S_obj = parents[1] if list.member.fixed_size(): idx = 1 if switch_obj is not None else 0 f.write('.HP\n') f.write('%s *\\fB%s\\fP(%s);\n' % (field.c_field_type, field.c_accessor_name, params[idx][0])) create_link('%s' % field.c_accessor_name) f.write('.HP\n') f.write('int \\fB%s\\fP(const %s *\\fIreply\\fP);\n' % (field.c_length_name, c_type)) create_link('%s' % field.c_length_name) if field.type.member.is_simple: f.write('.HP\n') f.write('xcb_generic_iterator_t \\fB%s\\fP(const %s *\\fIreply\\fP);\n' % (field.c_end_name, c_type)) create_link('%s' % field.c_end_name) else: f.write('.HP\n') f.write('%s \\fB%s\\fP(const %s *\\fIreply\\fP);\n' % (field.c_iterator_type, field.c_iterator_name, c_type)) create_link('%s' % field.c_iterator_name) for field in self.reply.fields: if field.type.is_list and not field.type.fixed_size(): _c_accessors_list(self, field) elif field.prev_varsized_field is not None or not field.type.fixed_size(): _c_accessors_field(self, field) f.write('.br\n') # Re-enable hyphenation and adjusting to both sides f.write('.hy 1\n') # argument reference f.write('.SH REQUEST ARGUMENTS\n') f.write('.IP \\fI%s\\fP 1i\n' % 'conn') f.write('The XCB connection to X11.\n') for field in param_fields: f.write('.IP \\fI%s\\fP 1i\n' % (field.c_field_name)) printed_enum = False # XXX: hard-coded until we fix xproto.xml if base_func_name == 'xcb_change_gc' and field.c_field_name == 'value_mask': field.enum = 'GC' elif base_func_name == 'xcb_change_window_attributes' and field.c_field_name == 'value_mask': field.enum = 'CW' elif base_func_name == 'xcb_create_window' and field.c_field_name == 'value_mask': field.enum = 'CW' if hasattr(field, "enum") and field.enum: # XXX: why the 'xcb' prefix? key = ('xcb', field.enum) if key in enums: f.write('One of the following values:\n') f.write('.RS 1i\n') enum = enums[key] count = len(enum.values) for (enam, eval) in enum.values: count = count - 1 f.write('.IP \\fI%s\\fP 1i\n' % (_n(key + (enam,)).upper())) if hasattr(enum, "doc") and enum.doc and enam in enum.doc.fields: desc = re.sub(r'`([^`]+)`', r'\\fI\1\\fP', enum.doc.fields[enam]) f.write('%s\n' % desc) else: f.write('TODO: NOT YET DOCUMENTED.\n') f.write('.RE\n') f.write('.RS 1i\n') printed_enum = True if hasattr(self, "doc") and self.doc and field.field_name in self.doc.fields: desc = self.doc.fields[field.field_name] desc = re.sub(r'`([^`]+)`', r'\\fI\1\\fP', desc) if printed_enum: f.write('\n') f.write('%s\n' % desc) else: f.write('TODO: NOT YET DOCUMENTED.\n') if printed_enum: f.write('.RE\n') # Reply reference if not void: f.write('.SH REPLY FIELDS\n') # These fields are present in every reply: f.write('.IP \\fI%s\\fP 1i\n' % 'response_type') f.write(('The type of this reply, in this case \\fI%s\\fP. This field ' 'is also present in the \\fIxcb_generic_reply_t\\fP and can ' 'be used to tell replies apart from each other.\n') % _n(self.reply.name).upper()) f.write('.IP \\fI%s\\fP 1i\n' % 'sequence') f.write('The sequence number of the last request processed by the X11 server.\n') f.write('.IP \\fI%s\\fP 1i\n' % 'length') f.write('The length of the reply, in words (a word is 4 bytes).\n') for field in self.reply.fields: if (field.c_field_name in frozenset(['response_type', 'sequence', 'length']) or field.c_field_name.startswith('pad')): continue if field.type.is_list and not field.type.fixed_size(): continue elif field.prev_varsized_field is not None or not field.type.fixed_size(): continue f.write('.IP \\fI%s\\fP 1i\n' % (field.c_field_name)) printed_enum = False if hasattr(field, "enum") and field.enum: # XXX: why the 'xcb' prefix? key = ('xcb', field.enum) if key in enums: f.write('One of the following values:\n') f.write('.RS 1i\n') enum = enums[key] count = len(enum.values) for (enam, eval) in enum.values: count = count - 1 f.write('.IP \\fI%s\\fP 1i\n' % (_n(key + (enam,)).upper())) if enum.doc and enam in enum.doc.fields: desc = re.sub(r'`([^`]+)`', r'\\fI\1\\fP', enum.doc.fields[enam]) f.write('%s\n' % desc) else: f.write('TODO: NOT YET DOCUMENTED.\n') f.write('.RE\n') f.write('.RS 1i\n') printed_enum = True if hasattr(self.reply, "doc") and self.reply.doc and field.field_name in self.reply.doc.fields: desc = self.reply.doc.fields[field.field_name] desc = re.sub(r'`([^`]+)`', r'\\fI\1\\fP', desc) if printed_enum: f.write('\n') f.write('%s\n' % desc) else: f.write('TODO: NOT YET DOCUMENTED.\n') if printed_enum: f.write('.RE\n') # text description f.write('.SH DESCRIPTION\n') if hasattr(self, "doc") and self.doc and self.doc.description: desc = self.doc.description desc = re.sub(r'`([^`]+)`', r'\\fI\1\\fP', desc) lines = desc.split('\n') f.write('\n'.join(lines) + '\n') f.write('.SH RETURN VALUE\n') if void: f.write(('Returns an \\fIxcb_void_cookie_t\\fP. Errors (if any) ' 'have to be handled in the event loop.\n\nIf you want to ' 'handle errors directly with \\fIxcb_request_check\\fP ' 'instead, use \\fI%s_checked\\fP. See ' '\\fBxcb-requests(%s)\\fP for details.\n') % (base_func_name, section)) else: f.write(('Returns an \\fI%s\\fP. Errors have to be handled when ' 'calling the reply function \\fI%s\\fP.\n\nIf you want to ' 'handle errors in the event loop instead, use ' '\\fI%s_unchecked\\fP. See \\fBxcb-requests(%s)\\fP for ' 'details.\n') % (cookie_type, self.c_reply_name, base_func_name, section)) f.write('.SH ERRORS\n') if hasattr(self, "doc") and self.doc: for errtype, errtext in self.doc.errors.items(): f.write('.IP \\fI%s\\fP 1i\n' % (_t(('xcb', errtype, 'error')))) errtext = re.sub(r'`([^`]+)`', r'\\fI\1\\fP', errtext) f.write('%s\n' % (errtext)) if not hasattr(self, "doc") or not self.doc or len(self.doc.errors) == 0: f.write('This request does never generate any errors.\n') if hasattr(self, "doc") and self.doc and self.doc.example: f.write('.SH EXAMPLE\n') f.write('.nf\n') f.write('.sp\n') lines = self.doc.example.split('\n') f.write('\n'.join(lines) + '\n') f.write('.fi\n') f.write('.SH SEE ALSO\n') if hasattr(self, "doc") and self.doc: see = ['.BR %s (%s)' % ('xcb-requests', section)] if self.doc.example: see.append('.BR %s (%s)' % ('xcb-examples', section)) for seename, seetype in self.doc.see.items(): if seetype == 'program': see.append('.BR %s (1)' % seename) elif seetype == 'event': see.append('.BR %s (%s)' % (_t(('xcb', seename, 'event')), section)) elif seetype == 'request': see.append('.BR %s (%s)' % (_n(('xcb', seename)), section)) elif seetype == 'function': see.append('.BR %s (%s)' % (seename, section)) else: see.append('TODO: %s (type %s)' % (seename, seetype)) f.write(',\n'.join(see) + '\n') f.write('.SH AUTHOR\n') f.write('Generated from %s.xml. Contact xcb@lists.freedesktop.org for corrections and improvements.\n' % _ns.header) f.close() def _man_event(self, name): if manpaths: sys.stdout.write('man/%s.%s ' % (self.c_type, section)) # Our CWD is src/, so this will end up in src/man/ f = open('man/%s.%s' % (self.c_type, section), 'w') f.write('.TH %s %s "%s" "%s" "XCB Events"\n' % (self.c_type, section, center_footer, left_footer)) # Left-adjust instead of adjusting to both sides f.write('.ad l\n') f.write('.SH NAME\n') brief = self.doc.brief if hasattr(self, "doc") and self.doc else '' f.write('%s \\- %s\n' % (self.c_type, brief)) f.write('.SH SYNOPSIS\n') # Don't split words (hyphenate) f.write('.hy 0\n') f.write('.B #include \n' % _ns.header) f.write('.PP\n') f.write('.SS Event datastructure\n') f.write('.nf\n') f.write('.sp\n') f.write('typedef %s %s {\n' % (self.c_container, self.c_type)) struct_fields = [] maxtypelen = 0 for field in self.fields: if not field.type.fixed_size() and not self.is_switch and not self.is_union: continue if field.wire: struct_fields.append(field) for field in struct_fields: length = len(field.c_field_type) # account for '*' pointer_spec if not field.type.fixed_size(): length += 1 maxtypelen = max(maxtypelen, length) def _c_complex_field(self, field, space=''): if (field.type.fixed_size() or # in case of switch with switch children, don't make the field a pointer # necessary for unserialize to work (self.is_switch and field.type.is_switch)): spacing = ' ' * (maxtypelen - len(field.c_field_type)) f.write('%s %s%s \\fI%s\\fP%s;\n' % (space, field.c_field_type, spacing, field.c_field_name, field.c_subscript)) else: print >> sys.stderr, 'ERROR: New unhandled documentation case' if not self.is_switch: for field in struct_fields: _c_complex_field(self, field) else: for b in self.bitcases: space = '' if b.type.has_name: space = ' ' for field in b.type.fields: _c_complex_field(self, field, space) if b.type.has_name: print >> sys.stderr, 'ERROR: New unhandled documentation case' pass f.write('} \\fB%s\\fP;\n' % self.c_type) f.write('.fi\n') f.write('.br\n') # Re-enable hyphenation and adjusting to both sides f.write('.hy 1\n') # argument reference f.write('.SH EVENT FIELDS\n') f.write('.IP \\fI%s\\fP 1i\n' % 'response_type') f.write(('The type of this event, in this case \\fI%s\\fP. This field is ' 'also present in the \\fIxcb_generic_event_t\\fP and can be used ' 'to tell events apart from each other.\n') % _n(name).upper()) f.write('.IP \\fI%s\\fP 1i\n' % 'sequence') f.write('The sequence number of the last request processed by the X11 server.\n') if not self.is_switch: for field in struct_fields: # Skip the fields which every event has, we already documented # them (see above). if field.c_field_name in ('response_type', 'sequence'): continue if isinstance(field.type, PadType): continue f.write('.IP \\fI%s\\fP 1i\n' % (field.c_field_name)) if hasattr(self, "doc") and self.doc and field.field_name in self.doc.fields: desc = self.doc.fields[field.field_name] desc = re.sub(r'`([^`]+)`', r'\\fI\1\\fP', desc) f.write('%s\n' % desc) else: f.write('NOT YET DOCUMENTED.\n') # text description f.write('.SH DESCRIPTION\n') if hasattr(self, "doc") and self.doc and self.doc.description: desc = self.doc.description desc = re.sub(r'`([^`]+)`', r'\\fI\1\\fP', desc) lines = desc.split('\n') f.write('\n'.join(lines) + '\n') if hasattr(self, "doc") and self.doc and self.doc.example: f.write('.SH EXAMPLE\n') f.write('.nf\n') f.write('.sp\n') lines = self.doc.example.split('\n') f.write('\n'.join(lines) + '\n') f.write('.fi\n') f.write('.SH SEE ALSO\n') if hasattr(self, "doc") and self.doc: see = ['.BR %s (%s)' % ('xcb_generic_event_t', section)] if self.doc.example: see.append('.BR %s (%s)' % ('xcb-examples', section)) for seename, seetype in self.doc.see.items(): if seetype == 'program': see.append('.BR %s (1)' % seename) elif seetype == 'event': see.append('.BR %s (%s)' % (_t(('xcb', seename, 'event')), section)) elif seetype == 'request': see.append('.BR %s (%s)' % (_n(('xcb', seename)), section)) elif seetype == 'function': see.append('.BR %s (%s)' % (seename, section)) else: see.append('TODO: %s (type %s)' % (seename, seetype)) f.write(',\n'.join(see) + '\n') f.write('.SH AUTHOR\n') f.write('Generated from %s.xml. Contact xcb@lists.freedesktop.org for corrections and improvements.\n' % _ns.header) f.close() def c_request(self, name): ''' Exported function that handles request declarations. ''' _c_type_setup(self, name, ('request',)) if self.reply: # Cookie type declaration _c_cookie(self, name) # Opcode define _c_opcode(name, self.opcode) # Request structure declaration _c_complex(self) if self.reply: _c_type_setup(self.reply, name, ('reply',)) # Reply structure definition _c_complex(self.reply) # Request prototypes has_fds = _c_reply_has_fds(self.reply) _c_request_helper(self, name, self.c_cookie_type, False, True, False, has_fds) _c_request_helper(self, name, self.c_cookie_type, False, False, False, has_fds) if self.need_aux: _c_request_helper(self, name, self.c_cookie_type, False, True, True, has_fds) _c_request_helper(self, name, self.c_cookie_type, False, False, True, has_fds) # Reply accessors _c_accessors(self.reply, name + ('reply',), name) _c_reply(self, name) if has_fds: _c_reply_fds(self, name) else: # Request prototypes _c_request_helper(self, name, 'xcb_void_cookie_t', True, False) _c_request_helper(self, name, 'xcb_void_cookie_t', True, True) if self.need_aux: _c_request_helper(self, name, 'xcb_void_cookie_t', True, False, True) _c_request_helper(self, name, 'xcb_void_cookie_t', True, True, True) # We generate the manpage afterwards because _c_type_setup has been called. # TODO: what about aux helpers? cookie_type = self.c_cookie_type if self.reply else 'xcb_void_cookie_t' _man_request(self, name, cookie_type, not self.reply, False) def c_event(self, name): ''' Exported function that handles event declarations. ''' # The generic event structure xcb_ge_event_t has the full_sequence field # at the 32byte boundary. That's why we've to inject this field into GE # events while generating the structure for them. Otherwise we would read # garbage (the internal full_sequence) when accessing normal event fields # there. force_packed = False if hasattr(self, 'is_ge_event') and self.is_ge_event and self.name == name: event_size = 0 for field in self.fields: if field.type.size != None and field.type.nmemb != None: event_size += field.type.size * field.type.nmemb if event_size == 32: full_sequence = Field(tcard32, tcard32.name, 'full_sequence', False, True, True) idx = self.fields.index(field) self.fields.insert(idx + 1, full_sequence) # If the event contains any 64-bit extended fields, they need # to remain aligned on a 64-bit boundary. Adding full_sequence # would normally break that; force the struct to be packed. force_packed = any(f.type.size == 8 and f.type.is_simple for f in self.fields[(idx+1):]) break _c_type_setup(self, name, ('event',)) # Opcode define _c_opcode(name, self.opcodes[name]) if self.name == name: # Structure definition _c_complex(self, force_packed) else: # Typedef _h('') _h('typedef %s %s;', _t(self.name + ('event',)), _t(name + ('event',))) _man_event(self, name) def c_error(self, name): ''' Exported function that handles error declarations. ''' _c_type_setup(self, name, ('error',)) # Opcode define _c_opcode(name, self.opcodes[name]) if self.name == name: # Structure definition _c_complex(self) else: # Typedef _h('') _h('typedef %s %s;', _t(self.name + ('error',)), _t(name + ('error',))) # Main routine starts here # Must create an "output" dictionary before any xcbgen imports. output = {'open' : c_open, 'close' : c_close, 'simple' : c_simple, 'enum' : c_enum, 'struct' : c_struct, 'union' : c_union, 'request' : c_request, 'event' : c_event, 'error' : c_error, } # Boilerplate below this point # Check for the argument that specifies path to the xcbgen python package. try: opts, args = getopt.getopt(sys.argv[1:], 'c:l:s:p:m') except getopt.GetoptError as err: print(err) print('Usage: c_client.py -c center_footer -l left_footer -s section [-p path] file.xml') sys.exit(1) for (opt, arg) in opts: if opt == '-c': center_footer=arg if opt == '-l': left_footer=arg if opt == '-s': section=arg if opt == '-p': sys.path.insert(1, arg) elif opt == '-m': manpaths = True sys.stdout.write('man_MANS = ') # Import the module class try: from xcbgen.state import Module from xcbgen.xtypes import * except ImportError: print(''' Failed to load the xcbgen Python package! Make sure that xcb/proto installed it on your Python path. If not, you will need to create a .pth file or define $PYTHONPATH to extend the path. Refer to the README file in xcb/proto for more info. ''') raise # Ensure the man subdirectory exists try: os.mkdir('man') except OSError as e: if e.errno != errno.EEXIST: raise # Parse the xml header module = Module(args[0], output) # Build type-registry and resolve type dependencies module.register() module.resolve() # Output the code module.generate()