| 1 | #!/usr/bin/env python2
|
| 2 | from __future__ import print_function
|
| 3 | from typing import List, Any, Dict, Iterator, Tuple, Optional, IO
|
| 4 | """help_gen.py
|
| 5 |
|
| 6 | Ideas for HTML -> ANSI converter:
|
| 7 |
|
| 8 | - `ls` -> <code>ls</code> -> is reverse video?
|
| 9 | - [link]() -> <a href=""> -> underlined, and then add a number to the bottom?
|
| 10 | - could also be bright blue
|
| 11 | - <pre> is also indented 4 spaces, like the markdown
|
| 12 | - red X <span class="X">X</span>
|
| 13 |
|
| 14 | - comments in code examples could be green?
|
| 15 |
|
| 16 | What about:
|
| 17 |
|
| 18 | - headings h1, h2, h3, h4
|
| 19 | - Right now cards use reverse video. Centering didn't look great.
|
| 20 |
|
| 21 | - <ul> - you could use a Unicode bullet here
|
| 22 | - <ol>
|
| 23 |
|
| 24 | Word wrapping? troff/groff doesn't do it, but they do this weird right-justify
|
| 25 | thing.
|
| 26 | """
|
| 27 |
|
| 28 | import cStringIO
|
| 29 | import HTMLParser
|
| 30 | import os
|
| 31 | import pprint
|
| 32 | import re
|
| 33 | import sys
|
| 34 |
|
| 35 | from typing import AnyStr
|
| 36 |
|
| 37 | from _devbuild.gen.htm8_asdl import h8_id
|
| 38 | from data_lang import htm8
|
| 39 | from doctools import html_lib
|
| 40 | from doctools.util import log
|
| 41 | from lazylex import html
|
| 42 |
|
| 43 | #from typing import List, Tuple
|
| 44 |
|
| 45 | # Sections have alphabetical characters, spaces, and '/' for I/O. They are
|
| 46 | # turned into anchors.
|
| 47 | SECTION_RE = re.compile(
|
| 48 | r'''
|
| 49 | \s*
|
| 50 | \[
|
| 51 | ([a-zA-Z0-9 /:-]+) # allow ysh:upgrade, byo-server-lib, etc.
|
| 52 | \]
|
| 53 | ''', re.VERBOSE)
|
| 54 |
|
| 55 | # Complex heuristic to highlight topics.
|
| 56 | TOPIC_RE = re.compile(
|
| 57 | r'''
|
| 58 | (X[ ])? # optional deprecation symbol X, then a single space
|
| 59 | @? # optional @array, e.g. @BASH_SOURCE
|
| 60 |
|
| 61 | ([a-zA-Z_][a-zA-Z0-9/:_-]+)
|
| 62 | # topic names: osh-usage, _status, ysh:all, BASH_REMATCH
|
| 63 | # List/append, cmd/append
|
| 64 |
|
| 65 | ( [ ] [^a-zA-Z0-9 ] \S*
|
| 66 | # trailer like >> or (make)
|
| 67 | |
|
| 68 | \(\) # optional () for func()
|
| 69 | )?
|
| 70 |
|
| 71 | ( # order of these 2 clauses matters
|
| 72 | [ ]*\n # spaces/newline
|
| 73 | |
|
| 74 | [ ]+ # 1 or more spaces
|
| 75 | )
|
| 76 | ''', re.VERBOSE)
|
| 77 | """
|
| 78 | ''', re.VERBOSE)
|
| 79 | """
|
| 80 |
|
| 81 |
|
| 82 | def _StringToHref(s):
|
| 83 | # lower case to match what doctools/cmark.py does
|
| 84 | return s.lower().replace(' ', '-')
|
| 85 |
|
| 86 |
|
| 87 | X_LEFT_SPAN = '<span style="color: darkred">'
|
| 88 |
|
| 89 |
|
| 90 | class TopicHtmlRenderer(object):
|
| 91 |
|
| 92 | def __init__(self, chapter, debug_out, linkify_stop_col):
|
| 93 | # type: (str, List, int) -> None
|
| 94 | self.chapter = chapter
|
| 95 | self.debug_out = debug_out
|
| 96 | self.linkify_stop_col = linkify_stop_col
|
| 97 |
|
| 98 | self.html_page = 'chap-%s.html' % chapter
|
| 99 |
|
| 100 | def _PrintTopic(self, m, out, line_info):
|
| 101 | # type: (Any, htm8.Output, Dict[str, Any]) -> None
|
| 102 | # The X
|
| 103 | topic_impl = True
|
| 104 | if m.group(1):
|
| 105 | out.PrintUntil(m.start(1))
|
| 106 | out.Print(X_LEFT_SPAN)
|
| 107 | out.PrintUntil(m.end(1))
|
| 108 | out.Print('</span>')
|
| 109 | topic_impl = False
|
| 110 |
|
| 111 | # The topic name to link
|
| 112 | topic = m.group(2)
|
| 113 | line_info['topics'].append((topic, topic_impl))
|
| 114 |
|
| 115 | out.PrintUntil(m.start(2))
|
| 116 | out.Print('<a href="%s#%s">' % (self.html_page, topic))
|
| 117 | out.PrintUntil(m.end(2))
|
| 118 | out.Print('</a>')
|
| 119 |
|
| 120 | def Render(self, line):
|
| 121 | # type: (str) -> str
|
| 122 | """Convert a line of text to HTML.
|
| 123 |
|
| 124 | Topics are highlighted and X made red.
|
| 125 |
|
| 126 | Args:
|
| 127 | chapter: where to link to
|
| 128 | line: RAW SPAN of HTML that is already escaped.
|
| 129 | debug_out: structured data
|
| 130 |
|
| 131 | Returns:
|
| 132 | The HTML with some tags inserted.
|
| 133 | """
|
| 134 | f = cStringIO.StringIO()
|
| 135 | out = htm8.Output(line, f)
|
| 136 |
|
| 137 | pos = 0 # position within line
|
| 138 |
|
| 139 | section_impl = True
|
| 140 |
|
| 141 | if line.startswith('X '):
|
| 142 | out.Print(X_LEFT_SPAN)
|
| 143 | out.PrintUntil(2)
|
| 144 | out.Print('</span>')
|
| 145 | pos = 2
|
| 146 | section_impl = False
|
| 147 | elif line.startswith(' '):
|
| 148 | pos = 2
|
| 149 | else:
|
| 150 | return line
|
| 151 |
|
| 152 | # Highlight [Section] at the start of a line.
|
| 153 | m = SECTION_RE.match(line, pos)
|
| 154 | if m:
|
| 155 | section_name = m.group(1)
|
| 156 | #href = _StringToHref(section_name)
|
| 157 | href = html_lib.PrettyHref(section_name, preserve_anchor_case=True)
|
| 158 |
|
| 159 | out.PrintUntil(m.start(1))
|
| 160 | out.Print('<a href="%s#%s" class="level2">' %
|
| 161 | (self.html_page, href))
|
| 162 | out.PrintUntil(m.end(1)) # anchor
|
| 163 | out.Print('</a>')
|
| 164 |
|
| 165 | pos = m.end(0) # ADVANCE
|
| 166 | else:
|
| 167 | section_name = None
|
| 168 |
|
| 169 | line_info = {
|
| 170 | 'section': section_name,
|
| 171 | 'impl': section_impl,
|
| 172 | 'topics': []
|
| 173 | }
|
| 174 | self.debug_out.append(line_info)
|
| 175 |
|
| 176 | # Whitespace after section, or leading whitespace
|
| 177 | _SPACE_1 = re.compile(r'[ ]+')
|
| 178 | m = _SPACE_1.match(line, pos)
|
| 179 | assert m, 'Expected whitespace %r' % line
|
| 180 |
|
| 181 | pos = m.end()
|
| 182 |
|
| 183 | # Keep matching topics until it doesn't match.
|
| 184 | while True:
|
| 185 | m = TOPIC_RE.match(line, pos)
|
| 186 |
|
| 187 | if not m:
|
| 188 | break
|
| 189 |
|
| 190 | pos = m.end()
|
| 191 |
|
| 192 | # The 1-based column number of the end of this topic
|
| 193 | col = m.end(2) + 1
|
| 194 | if self.linkify_stop_col != -1 and col > self.linkify_stop_col:
|
| 195 | #log('STOPPING %d > %d' % (col, self.linkify_stop_col))
|
| 196 | break
|
| 197 |
|
| 198 | self._PrintTopic(m, out, line_info)
|
| 199 |
|
| 200 | #log('trailing %r', line[pos:])
|
| 201 |
|
| 202 | out.PrintTheRest()
|
| 203 | return f.getvalue()
|
| 204 |
|
| 205 |
|
| 206 | class Splitter(HTMLParser.HTMLParser):
|
| 207 | """Split an HTML stream starting at each of the heading tags.
|
| 208 |
|
| 209 | For *-help.html.
|
| 210 |
|
| 211 | TODO: Rewrite with this with lazylex!
|
| 212 |
|
| 213 | Algorithm:
|
| 214 | - ExtractBody() first, then match balanced tags
|
| 215 | - SPLIT by h2, h3, h4
|
| 216 | - Match <pre><code> blocks and re-indent
|
| 217 | - Later:
|
| 218 | - links <a href="">
|
| 219 | - `` is turned into inline <code></code>
|
| 220 | - ** ** for bold
|
| 221 | - * * for emphasis
|
| 222 | - <p> needs word wrapping! Oops.
|
| 223 | - actually cmark seems to preserve this? OK maybe not.
|
| 224 | - we just need space between <p>
|
| 225 | """
|
| 226 |
|
| 227 | def __init__(self, heading_tags, out):
|
| 228 | # type: (List[str], List) -> None
|
| 229 | HTMLParser.HTMLParser.__init__(self)
|
| 230 | self.heading_tags = heading_tags
|
| 231 | self.out = out
|
| 232 |
|
| 233 | self.cur_group = None # type-not-checked: List[Tuple[str, str, List, List]]
|
| 234 | self.in_heading = False
|
| 235 |
|
| 236 | self.indent = 0
|
| 237 |
|
| 238 | def log(self, msg, *args):
|
| 239 | # type: (str, *Any) -> None
|
| 240 | ind = self.indent * ' '
|
| 241 | if 0:
|
| 242 | log(ind + msg, *args)
|
| 243 |
|
| 244 | def handle_starttag(self, tag, attrs):
|
| 245 | # type: (AnyStr, List[Tuple[AnyStr, AnyStr]]) -> None
|
| 246 | if tag in self.heading_tags:
|
| 247 | self.in_heading = True
|
| 248 | if self.cur_group:
|
| 249 | self.out.append(self.cur_group)
|
| 250 |
|
| 251 | self.cur_group = (tag, attrs, [], [])
|
| 252 |
|
| 253 | self.log('[%d] <> %s %s', self.indent, tag, attrs)
|
| 254 | self.indent += 1
|
| 255 |
|
| 256 | def handle_endtag(self, tag):
|
| 257 | # type: (AnyStr) -> None
|
| 258 | if tag in self.heading_tags:
|
| 259 | self.in_heading = False
|
| 260 |
|
| 261 | self.log('[%d] </> %s', self.indent, tag)
|
| 262 | self.indent -= 1
|
| 263 |
|
| 264 | def handle_entityref(self, name):
|
| 265 | # type: (AnyStr) -> None
|
| 266 | """
|
| 267 | From Python docs:
|
| 268 | This method is called to process a named character reference of the form
|
| 269 | &name; (e.g. >), where name is a general entity reference (e.g. 'gt').
|
| 270 | """
|
| 271 | c = html.CHAR_ENTITY[name]
|
| 272 | if self.in_heading:
|
| 273 | self.cur_group[2].append(c)
|
| 274 | else:
|
| 275 | if self.cur_group:
|
| 276 | self.cur_group[3].append(c)
|
| 277 |
|
| 278 | def handle_data(self, data):
|
| 279 | # type: (AnyStr) -> None
|
| 280 | self.log('data %r', data)
|
| 281 | if self.in_heading:
|
| 282 | self.cur_group[2].append(data)
|
| 283 | else:
|
| 284 | if self.cur_group:
|
| 285 | self.cur_group[3].append(data)
|
| 286 |
|
| 287 | def end(self):
|
| 288 | # type: () -> None
|
| 289 | if self.cur_group:
|
| 290 | self.out.append(self.cur_group)
|
| 291 |
|
| 292 | # Maybe detect nesting?
|
| 293 | if self.indent != 0:
|
| 294 | raise RuntimeError(
|
| 295 | 'Unbalanced HTML tags: indent=%d, cur_group=%s' %
|
| 296 | (self.indent, self.cur_group))
|
| 297 |
|
| 298 |
|
| 299 | def ExtractBody(s):
|
| 300 | # type: (str) -> str
|
| 301 | """Extract what's in between <body></body>
|
| 302 |
|
| 303 | The splitter needs balanced tags, and what's in <head> isn't
|
| 304 | balanced.
|
| 305 | """
|
| 306 | f = cStringIO.StringIO()
|
| 307 | out = htm8.Output(s, f)
|
| 308 | tag_lexer = htm8.TagLexer(s)
|
| 309 |
|
| 310 | pos = 0
|
| 311 | it = html.ValidTokens(s)
|
| 312 | while True:
|
| 313 | try:
|
| 314 | tok_id, end_pos = next(it)
|
| 315 | except StopIteration:
|
| 316 | break
|
| 317 |
|
| 318 | if tok_id == h8_id.StartTag:
|
| 319 | tag_lexer.Reset(pos, end_pos)
|
| 320 | if tag_lexer.TagName() == 'body':
|
| 321 | body_start_right = end_pos # right after <body>
|
| 322 |
|
| 323 | out.SkipTo(body_start_right)
|
| 324 | body_end_left, _ = html.ReadUntilEndTag(it, tag_lexer, 'body')
|
| 325 |
|
| 326 | out.PrintUntil(body_end_left)
|
| 327 | break
|
| 328 |
|
| 329 | pos = end_pos
|
| 330 |
|
| 331 | return f.getvalue()
|
| 332 |
|
| 333 |
|
| 334 | def SplitIntoCards(heading_tags, contents):
|
| 335 | # type: (List[str], str) -> Iterator
|
| 336 | contents = ExtractBody(contents)
|
| 337 |
|
| 338 | groups = []
|
| 339 | sp = Splitter(heading_tags, groups)
|
| 340 | sp.feed(contents)
|
| 341 | sp.end()
|
| 342 |
|
| 343 | for tag, attrs, heading_parts, parts in groups:
|
| 344 | heading = ''.join(heading_parts).strip()
|
| 345 |
|
| 346 | # Don't strip leading space?
|
| 347 | text = ''.join(parts)
|
| 348 | text = text.strip('\n') + '\n'
|
| 349 |
|
| 350 | #log('text = %r', text[:10])
|
| 351 |
|
| 352 | yield tag, attrs, heading, text
|
| 353 |
|
| 354 | #log('make_help.py: Parsed %d parts', len(groups))
|
| 355 |
|
| 356 |
|
| 357 | def HelpTopics(s):
|
| 358 | """
|
| 359 | Given a rendered toc-{osh,ysh}.html
|
| 360 |
|
| 361 | yield groups (section_id, section_name, block of text)
|
| 362 | """
|
| 363 | tag_lexer = htm8.TagLexer(s)
|
| 364 |
|
| 365 | pos = 0
|
| 366 | it = html.ValidTokens(s)
|
| 367 | while True:
|
| 368 | try:
|
| 369 | tok_id, end_pos = next(it)
|
| 370 | except StopIteration:
|
| 371 | break
|
| 372 |
|
| 373 | if tok_id == h8_id.StartTag:
|
| 374 | tag_lexer.Reset(pos, end_pos)
|
| 375 | #log('%r', tag_lexer.TagString())
|
| 376 | #log('%r', tag_lexer.TagName())
|
| 377 |
|
| 378 | # Capture <h2 id="foo"> first
|
| 379 | if tag_lexer.TagName() == 'h2':
|
| 380 | h2_start_right = end_pos
|
| 381 |
|
| 382 | open_tag_right = end_pos
|
| 383 | section_id = tag_lexer.GetAttrRaw('id')
|
| 384 | assert section_id, 'Expected id= in %r' % tag_lexer.TagString()
|
| 385 |
|
| 386 | h2_end_left, _ = html.ReadUntilEndTag(it, tag_lexer, 'h2')
|
| 387 |
|
| 388 | anchor_html = s[h2_start_right:h2_end_left]
|
| 389 | paren_pos = anchor_html.find('<') # remove HTML link
|
| 390 | if paren_pos == -1:
|
| 391 | section_name = anchor_html
|
| 392 | else:
|
| 393 | section_name = anchor_html[:paren_pos].strip()
|
| 394 |
|
| 395 | # Now find the <code></code> span
|
| 396 | _, code_start_right = html.ReadUntilStartTag(
|
| 397 | it, tag_lexer, 'code')
|
| 398 | css_class = tag_lexer.GetAttrRaw('class')
|
| 399 | assert css_class is not None
|
| 400 | assert css_class.startswith(
|
| 401 | 'language-chapter-links-'), tag_lexer.TagString()
|
| 402 |
|
| 403 | code_end_left, _ = html.ReadUntilEndTag(it, tag_lexer, 'code')
|
| 404 |
|
| 405 | text = html.ToText(s, code_start_right, code_end_left)
|
| 406 | yield section_id, section_name, text
|
| 407 |
|
| 408 | pos = end_pos
|
| 409 |
|
| 410 |
|
| 411 | class DocNode(object):
|
| 412 | """To visualize doc structure."""
|
| 413 |
|
| 414 | def __init__(self, name, attrs=None, text=None):
|
| 415 | # type: (str, Optional[List[Tuple[str, str]]], Optional[str]) -> None
|
| 416 | self.name = name
|
| 417 | self.attrs = attrs # for h2 and h3 links
|
| 418 | self.text = text
|
| 419 | self.children = []
|
| 420 |
|
| 421 |
|
| 422 | def CardsFromIndex(sh, out_prefix):
|
| 423 | sections = []
|
| 424 | for section_id, section_name, text in HelpTopics(sys.stdin.read()):
|
| 425 | if 0:
|
| 426 | log('section_id = %r', section_id)
|
| 427 | log('section_name = %r', section_name)
|
| 428 | log('')
|
| 429 | #log('text = %r', text[:20])
|
| 430 |
|
| 431 | topic = '%s-%s' % (sh, section_id) # e.g. ysh-overview
|
| 432 |
|
| 433 | path = os.path.join(out_prefix, topic)
|
| 434 | with open(path, 'w') as f:
|
| 435 | f.write('%s\n\n' %
|
| 436 | section_name) # section_id is printed dynamically
|
| 437 | f.write(text)
|
| 438 | #f.write('\n') # extra
|
| 439 | #log(' Wrote %s', path)
|
| 440 | sections.append(section_id)
|
| 441 |
|
| 442 | log(' (doctools/make_help) -> %d sections -> %s', len(sections),
|
| 443 | out_prefix)
|
| 444 |
|
| 445 |
|
| 446 | def CardsFromChapters(
|
| 447 | out_dir, # type: str
|
| 448 | tag_level, # type: str
|
| 449 | paths, # type: List[str]
|
| 450 | ):
|
| 451 | # type: (...) -> Tuple[Dict[str, Optional[str]], DocNode]
|
| 452 | """
|
| 453 | Args:
|
| 454 | paths: list of chap-*.html to read
|
| 455 | """
|
| 456 | topic_to_chap = {}
|
| 457 |
|
| 458 | root_node = DocNode('/')
|
| 459 | cur_h2_node = None
|
| 460 |
|
| 461 | for path in paths:
|
| 462 | with open(path) as f:
|
| 463 | contents = f.read()
|
| 464 |
|
| 465 | filename = os.path.basename(path)
|
| 466 |
|
| 467 | tmp, _ = os.path.splitext(filename)
|
| 468 | assert tmp.startswith('chap-')
|
| 469 | chapter_name = tmp[len('chap-'):]
|
| 470 |
|
| 471 | page_node = DocNode(filename)
|
| 472 |
|
| 473 | cards = SplitIntoCards(['h2', 'h3', 'h4'], contents)
|
| 474 |
|
| 475 | for tag, attrs, heading, text in cards:
|
| 476 | values = [v for k, v in attrs if k == 'id']
|
| 477 | id_value = values[0] if len(values) == 1 else None
|
| 478 |
|
| 479 | topic_id = (id_value if id_value else html_lib.PrettyHref(
|
| 480 | heading, preserve_anchor_case=True))
|
| 481 |
|
| 482 | if tag == 'h2':
|
| 483 | h2 = DocNode(topic_id, attrs=attrs)
|
| 484 | page_node.children.append(h2)
|
| 485 | cur_h2_node = h2
|
| 486 | elif tag == 'h3':
|
| 487 | # attach text so we can see which topics have empty bodies
|
| 488 | h3 = DocNode(topic_id, attrs=attrs, text=text)
|
| 489 | cur_h2_node.children.append(h3)
|
| 490 |
|
| 491 | if tag != tag_level:
|
| 492 | continue # we only care about h3 now
|
| 493 |
|
| 494 | if 0:
|
| 495 | log('tag = %r', tag)
|
| 496 | log('topic_id = %r', topic_id)
|
| 497 | log('heading = %r', heading)
|
| 498 | log('text = %r', text[:20])
|
| 499 |
|
| 500 | embed = ('oils-embed', '1') in attrs
|
| 501 |
|
| 502 | if out_dir is not None and embed:
|
| 503 | # indices start with _
|
| 504 | path = os.path.join(out_dir, topic_id)
|
| 505 | with open(path, 'w') as f:
|
| 506 | f.write(text)
|
| 507 |
|
| 508 | # help builtin will show URL if there's a chapter name
|
| 509 | topic_to_chap[topic_id] = None if embed else chapter_name
|
| 510 |
|
| 511 | root_node.children.append(page_node)
|
| 512 |
|
| 513 | num_sections = sum(len(child.children) for child in root_node.children)
|
| 514 |
|
| 515 | log(
|
| 516 | '%d chapters -> (doctools/make_help) -> %d <h3> cards from %d <h2> sections to %s',
|
| 517 | len(paths), len(topic_to_chap), num_sections, out_dir)
|
| 518 |
|
| 519 | return topic_to_chap, root_node
|
| 520 |
|
| 521 |
|
| 522 | class StrPool(object):
|
| 523 |
|
| 524 | def __init__(self):
|
| 525 | # type: () -> None
|
| 526 | self.var_names = {}
|
| 527 | self.global_strs = []
|
| 528 | self.unique_id = 1
|
| 529 |
|
| 530 | def Add(self, s):
|
| 531 | # type: (str) -> None
|
| 532 | if s in self.var_names:
|
| 533 | return
|
| 534 |
|
| 535 | var_name = 'gStr%d' % self.unique_id
|
| 536 | self.unique_id += 1
|
| 537 |
|
| 538 | import json
|
| 539 | # Use JSON as approximation for C++ string
|
| 540 | self.global_strs.append('GLOBAL_STR(%s, %s)' %
|
| 541 | (var_name, json.dumps(s)))
|
| 542 |
|
| 543 | self.var_names[s] = var_name
|
| 544 |
|
| 545 |
|
| 546 | def WriteTopicDict(topic_dict, header_f, cc_f):
|
| 547 | # type: (Dict[str, Optional[str]], IO[bytes], IO[bytes]) -> None
|
| 548 | header_f.write('''
|
| 549 | #include "mycpp/runtime.h"
|
| 550 |
|
| 551 | namespace help_meta {
|
| 552 | Dict<BigStr*, BigStr*>* TopicMetadata();
|
| 553 | }
|
| 554 | ''')
|
| 555 |
|
| 556 | pool = StrPool()
|
| 557 |
|
| 558 | for k, v in topic_dict.iteritems():
|
| 559 | pool.Add(k)
|
| 560 | if v is not None:
|
| 561 | pool.Add(v)
|
| 562 | #log('%s %s', k, v)
|
| 563 |
|
| 564 | num_items = len(topic_dict)
|
| 565 | key_names = []
|
| 566 | val_names = []
|
| 567 |
|
| 568 | for k, v in topic_dict.iteritems():
|
| 569 | key_names.append(pool.var_names[k])
|
| 570 | if v is None:
|
| 571 | v_str = 'nullptr'
|
| 572 | else:
|
| 573 | v_str = pool.var_names[v]
|
| 574 | val_names.append(v_str)
|
| 575 |
|
| 576 | cc_f.write('''
|
| 577 | #include "mycpp/runtime.h"
|
| 578 |
|
| 579 | namespace help_meta {
|
| 580 |
|
| 581 | %s
|
| 582 |
|
| 583 | GLOBAL_DICT(gTopics, BigStr*, BigStr*, %d, {%s}, {%s});
|
| 584 |
|
| 585 | Dict<BigStr*, BigStr*>* TopicMetadata() {
|
| 586 | return gTopics;
|
| 587 | }
|
| 588 | }
|
| 589 | ''' % ('\n'.join(pool.global_strs), num_items, ' COMMA '.join(key_names),
|
| 590 | ' COMMA '.join(val_names)))
|
| 591 |
|
| 592 |
|
| 593 | def main(argv):
|
| 594 | # type: (List[str]) -> None
|
| 595 | action = argv[1]
|
| 596 |
|
| 597 | if action == 'cards-from-index':
|
| 598 | sh = argv[2] # osh or ysh
|
| 599 | out_prefix = argv[3]
|
| 600 |
|
| 601 | # Read HTML from stdin
|
| 602 | # TODO: could pass a list of files to speed it up
|
| 603 | CardsFromIndex(sh, out_prefix)
|
| 604 |
|
| 605 | elif action == 'cards-from-chapters':
|
| 606 |
|
| 607 | out_dir = argv[2]
|
| 608 | py_out = argv[3]
|
| 609 | cc_prefix = argv[4]
|
| 610 | pages = argv[5:]
|
| 611 |
|
| 612 | topic_to_chap, _ = CardsFromChapters(out_dir, 'h3', pages)
|
| 613 |
|
| 614 | # Write topic dict as Python and C++
|
| 615 |
|
| 616 | with open(py_out, 'w') as f:
|
| 617 | f.write('TOPICS = %s\n' % pprint.pformat(topic_to_chap))
|
| 618 |
|
| 619 | f.write('''
|
| 620 |
|
| 621 | from typing import Dict
|
| 622 |
|
| 623 | def TopicMetadata():
|
| 624 | # type: () -> Dict[str, str]
|
| 625 | return TOPICS
|
| 626 | ''')
|
| 627 |
|
| 628 | h_path = cc_prefix + '.h'
|
| 629 | cc_path = cc_prefix + '.cc'
|
| 630 |
|
| 631 | with open(h_path, 'w') as header_f:
|
| 632 | with open(cc_path, 'w') as cc_f:
|
| 633 | WriteTopicDict(topic_to_chap, header_f, cc_f)
|
| 634 |
|
| 635 | elif action == 'ref-check':
|
| 636 | from doctools import cmark
|
| 637 | from doctools import oils_doc
|
| 638 | from doctools import ref_check
|
| 639 |
|
| 640 | chapters = []
|
| 641 | all_toc_nodes = []
|
| 642 |
|
| 643 | for path in argv[2:]:
|
| 644 | filename = os.path.basename(path)
|
| 645 |
|
| 646 | if filename.endswith('.md'):
|
| 647 | assert filename.startswith('toc-'), path
|
| 648 |
|
| 649 | # First convert to HTML
|
| 650 | with open(path) as in_file:
|
| 651 | html = cmark.md2html(in_file.read())
|
| 652 |
|
| 653 | # Now highlight code, which # which gives debug output for the
|
| 654 | # language-chapter-links-*
|
| 655 |
|
| 656 | box_nodes = []
|
| 657 | html = oils_doc.HighlightCode(html, None, debug_out=box_nodes)
|
| 658 | all_toc_nodes.append({'toc': filename, 'boxes': box_nodes})
|
| 659 |
|
| 660 | elif filename.endswith('.html'):
|
| 661 | assert filename.startswith('chap-'), path
|
| 662 | chapters.append(path)
|
| 663 |
|
| 664 | else:
|
| 665 | raise RuntimeError('Expected toc-* or chap-*, got %r' %
|
| 666 | filename)
|
| 667 |
|
| 668 | topics, chap_tree = CardsFromChapters(None, 'h3', chapters)
|
| 669 |
|
| 670 | #log('%d chapters: %s', len(chapters), chapters[:5])
|
| 671 | #log('%d topics: %s', len(topics), topics.keys()[:10])
|
| 672 | log('')
|
| 673 |
|
| 674 | # Compare TOC vs. chapters
|
| 675 | ref_check.Check(all_toc_nodes, chap_tree)
|
| 676 |
|
| 677 | else:
|
| 678 | raise RuntimeError('Invalid action %r' % action)
|
| 679 |
|
| 680 |
|
| 681 | if __name__ == '__main__':
|
| 682 | try:
|
| 683 | main(sys.argv)
|
| 684 | except RuntimeError as e:
|
| 685 | print('FATAL: %s' % e, file=sys.stderr)
|
| 686 | sys.exit(1)
|