[99038] | 1 | # -*- coding: utf-8 -*-
|
---|
| 2 | # $Id: build_id_to_file_mapping.py 99497 2023-04-21 01:55:02Z vboxsync $
|
---|
| 3 |
|
---|
| 4 | """
|
---|
| 5 | Scans the given files (globbed) for topic id and stores records the filename
|
---|
| 6 | in the output file.
|
---|
| 7 |
|
---|
| 8 | This is used by add_file_to_id_only_references.py after converting man_V*.xml
|
---|
| 9 | refentry files to dita to correct links.
|
---|
| 10 | """
|
---|
| 11 |
|
---|
| 12 | __copyright__ = \
|
---|
| 13 | """
|
---|
| 14 | Copyright (C) 2023 Oracle and/or its affiliates.
|
---|
| 15 |
|
---|
| 16 | This file is part of VirtualBox base platform packages, as
|
---|
| 17 | available from https://www.virtualbox.org.
|
---|
| 18 |
|
---|
| 19 | This program is free software; you can redistribute it and/or
|
---|
| 20 | modify it under the terms of the GNU General Public License
|
---|
| 21 | as published by the Free Software Foundation, in version 3 of the
|
---|
| 22 | License.
|
---|
| 23 |
|
---|
| 24 | This program is distributed in the hope that it will be useful, but
|
---|
| 25 | WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
| 26 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
---|
| 27 | General Public License for more details.
|
---|
| 28 |
|
---|
| 29 | You should have received a copy of the GNU General Public License
|
---|
| 30 | along with this program; if not, see <https://www.gnu.org/licenses>.
|
---|
| 31 |
|
---|
| 32 | SPDX-License-Identifier: GPL-3.0-only
|
---|
| 33 | """
|
---|
| 34 | __version__ = "$Revision: 99497 $"
|
---|
| 35 |
|
---|
| 36 |
|
---|
| 37 | # Standard python imports.
|
---|
| 38 | import glob;
|
---|
| 39 | import os;
|
---|
| 40 | import re;
|
---|
| 41 | import sys;
|
---|
| 42 |
|
---|
| 43 |
|
---|
| 44 | g_oReDita = re.compile(r'<topic[^><]*\sid=("[^">]+"|\'[^\'>]+\')');
|
---|
| 45 |
|
---|
| 46 | def scanDitaFileForIds(dIdToFile, sContent, sFile):
|
---|
| 47 | """
|
---|
| 48 | Scans the given content of a .dita-file for topic IDs that can be referenced.
|
---|
| 49 | """
|
---|
| 50 | for oMatch in g_oReDita.finditer(sContent):
|
---|
| 51 | sId = oMatch.group(1)[1:-1];
|
---|
| 52 | if sId:
|
---|
| 53 | dIdToFile[sId] = sFile;
|
---|
| 54 |
|
---|
| 55 | g_oReRefentry = re.compile(r'<(refentry\b|refsect[12345]\b|cmdsynopsis\b)[^><]*\sid=("[^">]+"|\'[^\'>]+\')');
|
---|
| 56 |
|
---|
| 57 | def scanDocbookRefentryForIds(dIdToFile, sContent, sFile):
|
---|
| 58 | """
|
---|
| 59 | Scans the given content of a Docbook refentry file for topic IDs that can be referenced.
|
---|
| 60 | """
|
---|
| 61 | for oMatch in g_oReRefentry.finditer(sContent):
|
---|
| 62 | sId = oMatch.group(2)[1:-1];
|
---|
| 63 | if sId:
|
---|
[99497] | 64 | #dIdToFile[sId] = sFile;
|
---|
| 65 | dIdToFile[sId] = '%s.dita' % (sId,);
|
---|
[99038] | 66 |
|
---|
| 67 | def isDocbook(sContent):
|
---|
| 68 | """
|
---|
| 69 | Check if the file is a Docbook one.
|
---|
| 70 | """
|
---|
| 71 | return sContent.find('<refentry ') >= 0 and sContent.find('<refentryinfo>');
|
---|
| 72 |
|
---|
| 73 | def error(sMessage):
|
---|
| 74 | """ Reports an error. """
|
---|
| 75 | print('build_id_to_file_mapping.py: error: %s' % sMessage, file = sys.stderr);
|
---|
| 76 | return 1;
|
---|
| 77 |
|
---|
| 78 | def syntax(sMessage):
|
---|
| 79 | """ Reports a syntax error. """
|
---|
| 80 | print('build_id_to_file_mapping.py: syntax error: %s' % sMessage, file = sys.stderr);
|
---|
| 81 | return 2;
|
---|
| 82 |
|
---|
| 83 | def usage():
|
---|
| 84 | """ Reports usage. """
|
---|
| 85 | print('usage: build_id_to_file_mapping.py --output <map.db> file1.dita docbook2.xml wild*card.* [...]');
|
---|
| 86 | return 0;
|
---|
| 87 |
|
---|
| 88 | def main(asArgs):
|
---|
| 89 | """
|
---|
| 90 | C-like main function.
|
---|
| 91 | """
|
---|
| 92 | #
|
---|
| 93 | # Process arguments.
|
---|
| 94 | #
|
---|
| 95 | dIdToFile = {};
|
---|
| 96 | sOutput = None;
|
---|
| 97 | fEndOfArgs = False;
|
---|
| 98 | iArg = 1;
|
---|
| 99 | while iArg < len(asArgs):
|
---|
| 100 | sArg = asArgs[iArg];
|
---|
| 101 | if sArg[0] == '-' and not fEndOfArgs:
|
---|
| 102 | # Options.
|
---|
| 103 | if sArg == '--':
|
---|
| 104 | fEndOfArgs = True;
|
---|
| 105 | elif sArg in ('--help', '-h', '-?'):
|
---|
| 106 | return usage();
|
---|
| 107 | elif sArg in ('--version', '-V' ):
|
---|
| 108 | print(__version__[__version__.find(':') + 2:-2]);
|
---|
| 109 | elif sArg in ('--output', '-o'):
|
---|
| 110 | iArg += 1;
|
---|
| 111 | if iArg >= len(asArgs):
|
---|
| 112 | return syntax('Expected filename following "--output"!');
|
---|
| 113 | sOutput = asArgs[iArg];
|
---|
| 114 | else:
|
---|
| 115 | return syntax('Unknown option: %s' % (sArg,));
|
---|
| 116 | else:
|
---|
| 117 | # Input files.
|
---|
[99059] | 118 | if sArg[0] == '@':
|
---|
| 119 | with open(sArg[1:], 'r', encoding = 'utf-8') as oFile:
|
---|
| 120 | asFiles = oFile.read().split();
|
---|
| 121 | else:
|
---|
| 122 | asFiles = glob.glob(sArg);
|
---|
[99038] | 123 | if not asFiles:
|
---|
| 124 | return error('File not found: %s' % (sArg,));
|
---|
| 125 | for sFile in asFiles:
|
---|
| 126 | try:
|
---|
| 127 | with open(sFile, 'r', encoding = 'utf-8') as oFile:
|
---|
| 128 | sContent = oFile.read();
|
---|
| 129 | except Exception as oXcpt: # pylint: disable=broad-exception-caught
|
---|
| 130 | return error('Failed to open and read "%s": %s' % (sFile, oXcpt,));
|
---|
| 131 | if isDocbook(sContent):
|
---|
| 132 | scanDocbookRefentryForIds(dIdToFile, sContent, os.path.splitext(os.path.basename(sFile))[0] + '.dita');
|
---|
| 133 | else:
|
---|
| 134 | scanDitaFileForIds(dIdToFile, sContent, os.path.basename(sFile));
|
---|
| 135 | iArg += 1;
|
---|
| 136 |
|
---|
| 137 | # Dump the dictionary.
|
---|
| 138 | asDict = sorted(['%s=%s' % (sKey, sValue) for sKey, sValue in dIdToFile.items()]);
|
---|
| 139 | if sOutput is not None:
|
---|
| 140 | try:
|
---|
| 141 | with open(sOutput, 'w', encoding = 'utf-8') as oFile:
|
---|
| 142 | oFile.write('\n'.join(asDict));
|
---|
| 143 | except Exception as oXcpt: # pylint: disable=broad-exception-caught
|
---|
| 144 | return error('Failed to open and write "%s": %s' % (sFile, oXcpt,));
|
---|
| 145 | else:
|
---|
| 146 | sys.stdout.write('\n'.join(asDict));
|
---|
| 147 | return 0;
|
---|
| 148 |
|
---|
| 149 | if __name__ == "__main__":
|
---|
| 150 | sys.exit(main(sys.argv));
|
---|