File indexing completed on 2024-02-18 05:44:35

0001 # -*- coding: UTF-8 -*-
0002 
0003 """
0004 Resolve aggregate messages produced by C{msgcat}.
0005 
0006 Documented in C{doc/user/sieving.docbook}.
0007 
0008 @author: Chusslove Illich (Часлав Илић) <caslav.ilic@gmx.net>
0009 @license: GPLv3
0010 """
0011 
0012 # DESIGN NOTE:
0013 # If one of the messages is missing one of the parts that others have,
0014 # that part is silently not added to the aggregation -- there is no explicit
0015 # indicator to tell that it was missing.
0016 # PO file names need not be unique either (if collected from a directory tree),
0017 # so it is not possible to deduce this from file names; likewise for project ID.
0018 # This means that there is no way to reconstruct complete original messages,
0019 # so each part has to be resolved independently.
0020 
0021 import re
0022 
0023 from pology import _, n_
0024 from pology.header import Header
0025 from pology.message import Message
0026 from pology.report import report
0027 from pology.sieve import SieveError
0028 
0029 
0030 def setup_sieve (p):
0031 
0032     p.set_desc(_("@info sieve discription",
0033     "Resolve aggregate messages produced by '%(cmd)s'.",
0034     cmd="msgcat"
0035     ))
0036 
0037     p.add_param("first", bool, defval=False,
0038                 desc=_("@info sieve parameter discription",
0039     "Always pick the first variant (by default, aggregate messages "
0040     "are resolved by taking the most frequent variant)."
0041     ))
0042     p.add_param("unfuzzy", bool, defval=False,
0043                 desc=_("@info sieve parameter discription",
0044     "Unfuzzy resolved messages. "
0045     "DANGEROUS: Use only if all messages in aggregation can be guaranteed "
0046     "not to be fuzzy."
0047     ))
0048     p.add_param("keepsrc", bool, defval=False,
0049                 desc=_("@info sieve parameter discription",
0050     "Keep source reference on resolved messages instead of removing them."
0051     ))
0052 
0053 
0054 class Sieve (object):
0055 
0056     def __init__ (self, params):
0057 
0058         exclusive_picks = [params.first]
0059         if sum(exclusive_picks) > 2:
0060             raise SieveError(
0061                 _("@info",
0062                   "Only one resolution criterion for "
0063                   "aggregate messages can be given."))
0064 
0065         if params.first:
0066             self.selvar = _selvar_first
0067         else:
0068             self.selvar = _selvar_frequent
0069 
0070         self.unfuzzy = params.unfuzzy
0071         self.keepsrc = params.keepsrc
0072 
0073         self.nresolved = 0
0074         self.nresolvedhdr = 0
0075 
0076 
0077     def process_header (self, hdr, cat):
0078 
0079         hmsg = Message(hdr.to_msg())
0080         if _resolve_msg(hmsg, self.selvar):
0081             self.nresolvedhdr += 1
0082         cat.header = Header(hmsg)
0083 
0084 
0085     def process (self, msg, cat):
0086 
0087         if _resolve_msg(msg, self.selvar):
0088             self.nresolved += 1
0089             if self.unfuzzy:
0090                 msg.unfuzzy()
0091             if not self.keepsrc:
0092                 msg.source[:] = []
0093 
0094 
0095     def finalize (self):
0096 
0097         if self.nresolvedhdr > 0:
0098             msg = n_("@info:progress",
0099                      "Resolved %(num)d aggregate header.",
0100                      "Resolved %(num)d aggregate headers.",
0101                      num=self.nresolvedhdr)
0102             report("===== " + msg)
0103         if self.nresolved > 0:
0104             msg = n_("@info:progress",
0105                      "Resolved %(num)d aggregate message.",
0106                      "Resolved %(num)d aggregate messages.",
0107                      num=self.nresolved)
0108             report("===== " + msg)
0109 
0110 
0111 def _selvar_first (texts):
0112 
0113     return texts[0]
0114 
0115 
0116 def _selvar_frequent (texts):
0117 
0118     tinds_by_text = {}
0119     for text, tind in zip(texts, list(range(len(texts)))):
0120         if text not in tinds_by_text:
0121             tinds_by_text[text] = []
0122         tinds_by_text[text].append(tind)
0123     tinds = sorted(list(tinds_by_text.values()), key=lambda x: (-len(x), x))
0124 
0125     return texts[tinds[0][0]]
0126 
0127 
0128 def _resolve_msg (msg, selvar):
0129 
0130     oldcount = msg.modcount
0131 
0132     if msg.manual_comment:
0133         aggtext = "\n".join(msg.manual_comment)
0134         msg.manual_comment[:] = _resolve_aggtext(aggtext, selvar).split("\n")
0135 
0136     if msg.auto_comment:
0137         aggtext = "\n".join(msg.auto_comment)
0138         msg.auto_comment[:] = _resolve_aggtext(aggtext, selvar).split("\n")
0139 
0140     # Separator swallows trailing newline, put it based on msgid.
0141     need_trailing_nl = msg.msgid.endswith("\n")
0142     for i in range(len(msg.msgstr)):
0143         nmsgstr = _resolve_aggtext(msg.msgstr[i], selvar)
0144         if need_trailing_nl and nmsgstr != msg.msgstr[i]:
0145             nmsgstr += "\n"
0146         msg.msgstr[i] = nmsgstr
0147 
0148     return msg.modcount > oldcount
0149 
0150 
0151 _splitter_rx = re.compile(r"\n?(?:#-){3,}# .*? (?:#-){3,}#\n?")
0152 
0153 def _resolve_aggtext (aggtext, selvar):
0154 
0155     texts = _splitter_rx.split(aggtext)[1:]
0156     return str(selvar(texts)) if texts else aggtext
0157