summaryrefslogtreecommitdiff
path: root/scripts/code_tools/modified_strings.py
blob: c777fc8c0de471070418b73518ea3e5f68bcc4c7 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
#!/usr/bin/env python3
"""\

This script scans the SL codebase for translation-related strings.

$LicenseInfo:firstyear=2020&license=viewerlgpl$
Second Life Viewer Source Code
Copyright (C) 2020, Linden Research, Inc.

This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation;
version 2.1 of the License only.

This library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
Lesser General Public License for more details.

You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA

Linden Research, Inc., 945 Battery Street, San Francisco, CA  94111  USA
$/LicenseInfo$
"""



import xml.etree.ElementTree as ET
import argparse
import os
import sys
from git import Repo, Git # requires the gitpython package
import pandas as pd
import re
from datetime import datetime

usage_msg="""%(prog)s [options]

Analyze the XUI configuration files to find text that may need to
be translated. Works by comparing two specified revisions, one
specified by --rev (default HEAD) and one specified by --rev_base
(default master). The script works by comparing xui contents of the
two revisions, and outputs a spreadsheet listing any areas of
difference. The target language must be specified using the --lang
option. Output is an excel file, which can be used as-is or imported
into google sheets.

If the --rev revision already contains a translation for the text, it
will be included in the spreadsheet for reference.

Normally you would want --rev_base to be the last revision to have
translations added, and --rev to be the tip of the current
project. You can find the last commit with translation work using "git log --grep INTL- | head"

The --missing argument can be used to find all text with missing
translations, regardless of when it was added. If translations are being kept
reasonably current, you will normally not need this argument.
"""

translate_attribs = [
    "title",
    "short_title",
    "value",
    "label",
    "label_selected",
    "tool_tip",
    "ignoretext",
    "yestext",
    "notext",
    "canceltext",
    "description",
    "longdescription"
]

def codify_for_print(val):
    if isinstance(val, str):
        return val.encode("utf-8")
    else:
        return str(val, 'utf-8').encode("utf-8")

# Returns a dict of { name => xml_node }
def read_xml_elements(blob):
    try:
        contents = blob.data_stream.read()
    except:
        # default - pretend we read a file with no elements of interest.
        # Parser will complain if it gets no elements at all.
        contents = '<?xml version="1.0" encoding="utf-8" standalone="yes" ?><strings></strings>'
    xml = ET.fromstring(contents)
    elts = {}
    for child in xml.iter():
        if "name" in child.attrib:
            name = child.attrib['name']
            elts[name] = child
    return elts

def failure(*msg):
    print(*msg)
    sys.exit(1)

# return True iff any element of lis is "in" thing
def has_any(thing,lis):
    for l in lis:
        if l in thing:
            return True
    return False

def should_translate(filename, elt, field, val):
    if val is None:
        return False
    # Should translate apply recursively?
    if "translate" in elt.attrib and elt.attrib["translate"] == "false":
        return False
    if has_any(filename,["floater_test","floater_aaa","floater_ui_preview"]):
        return False
    if "TestString PleaseIgnore" in val:
        return False
    val = re.sub(r"\[.*?\]","",val)
    if len(val) == 0:
        return False
    if val.isspace():
        return False
    val = val.strip()
    if val.isdigit():
        return False
    if not re.search('\w+', val):
        return False
    if re.match(r"^\s*\d*\s*x\s*\d*\s*$", val):
        #print(val, "matches resolution string, will ignore")
        return False
    # "value" attribute is a hairball, mostly used to encode non-display info but a few exceptions
    if field == "value":
        if elt.text is not None and len(elt.text) > 0:
            #print("value has text, ignoring", ET.tostring(elt))
            return False
        if has_any(elt.attrib,["label"]):
            return False
        if elt.tag in ["string","text"]:
            return True
        #print("including value attribute", val, "tag", elt.tag,"in", ET.tostring(elt))
        return True
    return True

def make_translation_table(mod_tree, base_tree, lang, args):

    xui_path = "{}/{}".format(xui_base, args.base_lang)
    try:
        mod_xui_tree = mod_tree[xui_path]
    except:
        failure("xui tree not found for base language", args.base_lang,"or target lang", lang)

    if args.rev == args.rev_base:
        failure("Revs are the same, nothing to compare")


    data = []
    # For all files to be checked for translations
    all_en_strings = set()
    for mod_blob in mod_xui_tree.traverse():
        filename = mod_blob.path
        if mod_blob.type == "tree": # directory, skip
            continue
        if args.files and os.path.basename(filename) not in args.files:
            continue # process only the specified files

        if args.verbose:
            print(filename)

        try:
            base_blob = base_tree[filename]
        except:
            if args.verbose:
                print("No matching base file found for", filename)
            base_blob = None

        try:
            transl_filename = filename.replace("/xui/{}/".format(args.base_lang), "/xui/{}/".format(lang))
            transl_blob = mod_tree[transl_filename]
        except:
            if args.verbose:
                print("No matching translation file found at", transl_filename)
            transl_blob = None

        mod_dict = read_xml_elements(mod_blob)
        base_dict = read_xml_elements(base_blob)
        transl_dict = read_xml_elements(transl_blob)

        rows = 0
        for name in list(mod_dict.keys()):
            if not name in base_dict or mod_dict[name].text != base_dict[name].text or (args.missing and not name in transl_dict):
                elt = mod_dict[name]
                val = elt.text
                field = "text"
                if should_translate(filename, elt, field, val):
                    transl_val = "--"
                    if name in transl_dict:
                        transl_val = transl_dict[name].text
                    if val in all_en_strings:
                        new_val = "(DUPLICATE)"
                    else:
                        new_val = ""
                    data.append([val, transl_val, new_val, "", "", filename, name, field])
                    all_en_strings.add(val)
                    rows += 1
            for attr in translate_attribs:
                if attr in mod_dict[name].attrib:
                    if name not in base_dict \
                    or attr not in base_dict[name].attrib \
                    or mod_dict[name].attrib[attr] != base_dict[name].attrib[attr] \
                    or (args.missing and (not name in transl_dict or not attr in transl_dict[name].attrib)):
                        elt = mod_dict[name]
                        val = elt.attrib[attr]
                        if should_translate(filename, elt, attr, val):
                            transl_val = "--"
                            if name in transl_dict and attr in transl_dict[name].attrib:
                                transl_val = transl_dict[name].attrib[attr]
                            if val in all_en_strings:
                                new_val = "(DUPLICATE)"
                            else:
                                new_val = ""
                            #attr = attr + ":" + ET.tostring(elt)
                            data.append([val, transl_val, new_val, "", "", filename, name, attr])
                            all_en_strings.add(val)
                            rows += 1

    return data

def find_deletions(mod_tree, base_tree, lang, args, f):

    transl_xui_path = "{}/{}".format(xui_base, lang)
    try:
        transl_xui_tree = mod_tree[transl_xui_path]
    except:
        failure("xui tree not found for base language", args.base_lang,"or target lang", lang)

    for transl_blob in transl_xui_tree.traverse():
        if transl_blob.type == "tree": # directory, skip
            continue
        transl_filename = transl_blob.path
        mod_filename = transl_filename.replace("/xui/{}/".format(lang), "/xui/{}/".format(args.base_lang))
        #print("checking",transl_filename,"against",mod_filename)
        try:
            mod_blob = mod_tree[mod_filename]
        except:
            print("  delete file", transl_filename, file=f)
            continue
        mod_dict = read_xml_elements(mod_blob)
        if len(mod_dict) == 0:
            print("  delete file", transl_filename, file=f)
            continue
        transl_dict = read_xml_elements(transl_blob)
        #print("mod vs transl", len(mod_dict), len(transl_dict))
        lines = 0
        for elt_key in transl_dict:
            if not elt_key in mod_dict:
                if lines == 0:
                    print("  in file", transl_filename, file=f)
                lines += 1
                print("    delete element", elt_key, file=f)
            else:
                transl_elt = transl_dict[elt_key]
                mod_elt = mod_dict[elt_key]
                for a in transl_elt.attrib:
                    if not a in mod_elt.attrib:
                        if lines == 0:
                            print("  in file", transl_filename, file=f)
                        lines += 1
                        print("    delete attribute", a, "from", elt_key, file=f)
                if transl_elt.text and (not mod_elt.text):
                    if lines == 0:
                        print("  in file", transl_filename, file=f)
                    lines += 1
                    print("    delete text from", elt_key, file=f)

def save_translation_file(per_lang_data, aux_data, outfile):

    langs = sorted(per_lang_data.keys())
    print("Saving languages", ",".join(langs),"as",outfile)

    writer = pd.ExcelWriter(outfile, engine='xlsxwriter')

    workbook = writer.book
    wrap_format = workbook.add_format({'text_wrap': True})
    bold_wrap_format = workbook.add_format({'text_wrap': True, 'bold': True})
    wrap_unlocked_format = workbook.add_format({'text_wrap': True, 'locked': False})

    for lang in langs:
        data = per_lang_data[lang]
        num_translations = len(data)
        cols = ["EN", "Previous Translation ({})".format(lang.upper()), "ENTER NEW TRANSLATION ({})".format(lang.upper()), "Translator Questions", "Notes", "File", "Element", "Field"]
        df = pd.DataFrame(data, columns=cols)
        df.to_excel(writer, index=False, sheet_name = lang.upper())

        worksheet = writer.sheets[lang.upper()]

        # Translators primarily care about columns A-C, and should write
        # only in column C. Hide the others. Set widths.
        worksheet.protect()
        worksheet.set_column('A:B', 60, wrap_format)
        worksheet.set_column('C:C', 60, wrap_unlocked_format)
        worksheet.set_column('D:E', 40, wrap_unlocked_format)
        worksheet.set_column('F:F', 50, wrap_format, {'hidden': True})
        worksheet.set_column('G:H', 30, wrap_format, {'hidden': True})

        # Lock the top row (column headers) in place while scrolling
        worksheet.freeze_panes(1, 0)
        print("Added", num_translations, "rows for language", lang)

    # Reference info, not for translation
    for aux, data in list(aux_data.items()):
        df = pd.DataFrame(data, columns = ["Key", "Value"])
        df.to_excel(writer, index=False, sheet_name=aux)
        worksheet = writer.sheets[aux]
        worksheet.set_column('A:A', 50, bold_wrap_format)
        worksheet.set_column('B:B', 80, wrap_format)

    print("Writing", outfile)
    writer.save()

if __name__ == "__main__":

    parser = argparse.ArgumentParser(description="analyze viewer xui files for needed translations", usage=usage_msg)
    parser.add_argument("-v","--verbose", action="store_true", help="verbose flag")
    parser.add_argument("--missing", action="store_true", default = False, help="include all fields for which a translation does not exist")
    parser.add_argument("--deleted", action="store_true", default = False, help="show all translated entities which don't exist in english")
    parser.add_argument("--skip_spreadsheet", action="store_true", default = False, help="skip creating the translation spreadsheet")
    parser.add_argument("--rev", help="revision with modified strings, default HEAD", default="HEAD")
    parser.add_argument("--rev_base", help="previous revision to compare against, default main", default="main")
    parser.add_argument("--base_lang", help="base language, default en (normally leave unchanged - other values are only useful for testing)", default="en")
    parser.add_argument("--lang", help="target languages, or 'all_valid' or 'supported'; default is 'supported'", nargs="+", default = ["supported"])
    parser.add_argument("--files", help='list of files to process', metavar='F', type=str, nargs='*')
    parser.add_argument("--outfile", help='name of the output file', type=str, nargs='?', default="SL_Translations.xlsx")
    args = parser.parse_args()

    cwd = os.getcwd()
    rootdir = Git(cwd).rev_parse("--show-toplevel")
    repo = Repo(rootdir)
    try:
        mod_commit = repo.commit(args.rev)
    except:
        failure(args.rev,"is not a valid commit")
    try:
        base_commit = repo.commit(args.rev_base)
    except:
        failure(args.rev_base,"is not a valid commit")

    print("Will identify changes in", args.rev, "not present in", args.rev_base)
    if args.missing:
        print("Will also include any text for which no corresponding translation exists, regardless of when it was added")
    sys.stdout.flush()

    mod_tree = mod_commit.tree
    base_tree = base_commit.tree

    xui_base = "indra/newview/skins/default/xui"
    xui_base_tree = mod_tree[xui_base]

    # Find target languages
    # all languages present in the codebase
    valid_langs = [tree.name.lower() for tree in xui_base_tree if tree.name.lower() != args.base_lang.lower()]
    # offically supported languages
    supported_langs = ["fr", "es", "it", "pt", "ja", "de"]
    langs = [l.lower() for l in args.lang]
    if "supported" in args.lang:
        langs = supported_langs
    if "all_valid" in args.lang:
        langs = valid_langs
    langs = sorted(langs)
    for lang in langs:
          if not lang in valid_langs:
              failure("Unknown target language {}. Valid values are {}".format(lang,", ".join(sorted(valid_langs) + ["all_valid","supported"])))
    print("Target language(s) are", ",".join(sorted(langs)))
    sys.stdout.flush()

    outfile = args.outfile
    try:
        f = open(outfile,"a+")
        f.close()
    except:
        failure("Can't write to output file",outfile,". Is it already open?")

    aux_data = { "REFERENCE": [["Command", " ".join(sys.argv)],
                               ["Date", str(datetime.now())],
                               ["Mod Commit", mod_commit.hexsha],
                               ["Base Commit", base_commit.hexsha],
                              ] }

    if not args.skip_spreadsheet:
        per_lang_data = {}
        for lang in langs:
            print("Creating spreadsheet for language", lang)
            sys.stdout.flush()

            per_lang_data[lang] = make_translation_table(mod_tree, base_tree, lang, args)

        print("Saving output file", outfile)
        save_translation_file(per_lang_data, aux_data, outfile)

    if args.deleted:
        deletion_file = "Translate_deletions.txt"
        print("Saving deletion info to", deletion_file)
        with open(deletion_file,"w") as f:
            for lang in langs:
                find_deletions(mod_tree, base_tree, lang, args, f)