mirror of
https://github.com/Abdess/retroarch_system.git
synced 2026-10-27 06:07:24 -05:00
feat: search fields and descriptions from pages
This commit is contained in:
Abdessamad Derraz committed
1 parent
a93e9ee406
commit
fcb4aa27da
6 files changed
+238
-49
No files matched your search
+141
-35
@@ -2109,12 +2109,23 @@ def _browser_title(relative: Path, title: str) -> str:
|
||||
}
|
||||
if key in index_titles:
|
||||
return index_titles[key]
|
||||
if key in PART_PAGES:
|
||||
# A console page is filed with the parts of its manufacturer's page, and
|
||||
# titled "Sega: Dreamcast". A part names its range of files.
|
||||
part = key in PART_PAGES and (
|
||||
not key.startswith("systems/") or ": " not in title
|
||||
or re.search(r", part \d+ of \d+$", title))
|
||||
if part:
|
||||
return title
|
||||
# The words a web search for a console or an emulator types: "dreamcast
|
||||
# bios", "duckstation bios". The h1 and the sidebar keep the name.
|
||||
if key.startswith("emulators/"):
|
||||
return f"{title} emulator firmware"
|
||||
has_files = PAGE_DATA.get(key, {}).get("hasPart")
|
||||
# Lexaloffle, QEMU and ScummVM name a system page too.
|
||||
own = " emulator" if f"systems/{relative.name}" in PAGE_DATA else ""
|
||||
return f"{title}{own} BIOS files" if has_files else f"{title} emulator"
|
||||
if key.startswith("systems/"):
|
||||
return f"{title} systems"
|
||||
maker, _, console = title.partition(": ")
|
||||
return f"{console} BIOS files ({maker})" if console else f"{title} BIOS files"
|
||||
return title
|
||||
|
||||
|
||||
@@ -2157,45 +2168,129 @@ def _page_description(markdown: str, relative: Path, title: str) -> str:
|
||||
}
|
||||
if key in specific:
|
||||
return specific[key]
|
||||
if key in PART_PAGES:
|
||||
return (f"{title}: every file with its hashes, size, source references "
|
||||
"and who reads it.")
|
||||
# What a search engine shows is the page's own first sentence: 707 pages
|
||||
# shared five templates of level 3 words ("source-pinned", "provenance").
|
||||
browser = _browser_title(relative, title)
|
||||
held = re.search(r"\bThis part holds (.+?)\.(?: |$)", markdown, re.M)
|
||||
if key in PART_PAGES and held:
|
||||
return _shorten(f"{browser}. This part holds {_plain_markdown(held[1])}.", 155)
|
||||
lead = _lead(markdown)
|
||||
if key in PART_PAGES and re.search(r"\bpart \d+ of \d+", title):
|
||||
# The parts of one list open on the same sentence.
|
||||
first = _first_paragraph(lead) or _first_paragraph(markdown)
|
||||
return _shorten(f"{title}. {first}", 155)
|
||||
if key.startswith(f"{release_record.GET_DIR}/"):
|
||||
return (f"{title}: the BIOS files of this setup, as the ZIP of the latest "
|
||||
"release or one command, at an address that stays the same.")
|
||||
if key.startswith("platforms/") and relative.stem != "index":
|
||||
return (f"{title}: where the BIOS files go, the command that installs them, the pack, "
|
||||
"and every file by system with its hashes.")
|
||||
if key.startswith("emulators/") and relative.stem != "index":
|
||||
# Pinned only when the profile names the revision it was read at:
|
||||
# 48 pages without one called themselves source-pinned.
|
||||
about = PAGE_DATA.get(key, {}).get("about", {})
|
||||
pinned = any(a.get("version") for a in (about if isinstance(about, list) else [about]))
|
||||
kind = "source-pinned emulator firmware profile" if pinned else "emulator firmware profile"
|
||||
if not PAGE_DATA.get(key, {}).get("hasPart"):
|
||||
# 152 cores that load no file announced paths and hashes.
|
||||
kind = kind.replace(" firmware", "")
|
||||
trees = (" The data directories it reads are listed."
|
||||
if "\n## Data directories\n" in markdown else "")
|
||||
return f"{title}: {kind}. No BIOS or firmware file is required.{trees}"
|
||||
return f"{title}: {kind} with paths, hashes, requirements and validation behavior."
|
||||
size = re.search(r"^- (.+ to download\b.*?)\.?$", markdown, re.M)
|
||||
if size:
|
||||
return _shorten(f"{title}: the BIOS pack of this setup, {_plain_markdown(size[1])}. "
|
||||
"The address stays the same and always gives the latest ZIP.", 155)
|
||||
return _shorten(_first_paragraph(lead) or title, 155)
|
||||
if key.startswith("systems/") and relative.stem != "index":
|
||||
return f"{title}: indexed firmware files, hashes, variants, provenance and platform or emulator usage."
|
||||
return _system_description(markdown, browser)
|
||||
if key.startswith("emulators/") and relative.stem != "index":
|
||||
return _emulator_description(markdown, key, title, _first_paragraph(lead))
|
||||
first = _first_paragraph(lead, minimum=35)
|
||||
questions = [_plain_markdown(re.sub(r"\s*\{[^}]*\}\s*$", "", heading))
|
||||
for heading in re.findall(r"^## (.+\?)(?:[ \t]*\{[^}]*\})?[ \t]*$", markdown, re.M)]
|
||||
if not first and len(questions) >= 3:
|
||||
# A page of questions with no intro: its first paragraph is the
|
||||
# answer to one of them ("Run the command with --dry-run").
|
||||
return _shorten(f"{title}: {' '.join(questions)}", 155)
|
||||
first = first or _first_paragraph(markdown, minimum=35)
|
||||
if first:
|
||||
return _shorten(first, 155)
|
||||
return f"{title}: RetroBIOS source-traced retrogaming firmware reference."
|
||||
|
||||
paragraphs = re.split(r"\n\s*\n", markdown)
|
||||
for paragraph in paragraphs:
|
||||
|
||||
def _lead(markdown: str) -> str:
|
||||
"""The text a page opens with, before its first section, details
|
||||
block or ignored section (the install box, the profile table)."""
|
||||
return re.split(r"^(?:## |<details|<section)", markdown, maxsplit=1, flags=re.M)[0]
|
||||
|
||||
|
||||
def _emulator_description(markdown: str, key: str, title: str, intro: str) -> str:
|
||||
"""An emulator page by its intro sentence ("DuckStation emulates Sony
|
||||
PlayStation. It picks the BIOS for your game's region: ..."), or by
|
||||
its profile table when the profile names no system."""
|
||||
notes = re.search(r"<summary>Technical notes</summary>\s*\n\n(.+?)(?:\n\n|$)", markdown, re.S)
|
||||
if not intro and notes:
|
||||
sentence = re.split(r"(?<=\.)\s", _plain_markdown(notes[1]), maxsplit=1)[0]
|
||||
intro = f"{title}: {sentence[:1].lower()}{sentence[1:]}"
|
||||
if not intro:
|
||||
row = {name: _plain_markdown(value) for name, value in re.findall(
|
||||
r"^\| (Type|Classification|Version) \| (.+?) \|$", markdown, re.M)}
|
||||
kind = row.get("Classification") or row.get("Type") or "emulator"
|
||||
version = row.get("Version", "")
|
||||
known = f", version {version}" if version and version[0].isdigit() else ""
|
||||
intro = f"{title}: {kind.lower()}{known}."
|
||||
elif title not in intro:
|
||||
# "It reads 1 file": the profile names no system to open on.
|
||||
intro = f"{title}: {intro[:1].lower()}{intro[1:]}"
|
||||
# An alias page sends the reader to the core whose files it shares.
|
||||
if PAGE_DATA.get(key, {}).get("hasPart") or "uses the same firmware as" in intro:
|
||||
return _shorten(intro, 155)
|
||||
needs = "It needs no BIOS or firmware file."
|
||||
if "\n## Data directories" in markdown:
|
||||
needs = "It needs no BIOS or firmware file, only the data folders listed."
|
||||
return _shorten(f"{intro} {needs}", 155)
|
||||
|
||||
|
||||
# Lines a page opens with that say where it sits, not what it holds.
|
||||
_PLACE_LINES = ("Part of ", "This part holds ", "On a card, ")
|
||||
|
||||
|
||||
def _first_paragraph(markdown: str, minimum: int = 1) -> str:
|
||||
"""The first paragraph of prose of a page body, as plain text: no
|
||||
heading, table, list, fence, admonition or HTML block, and none that
|
||||
only says where the page sits."""
|
||||
for paragraph in re.split(r"\n\s*\n", markdown):
|
||||
stripped = paragraph.strip()
|
||||
if (
|
||||
not stripped
|
||||
or stripped.startswith(("#", "|", "```", "???", ":::", "<", "- ", "* "))
|
||||
or stripped.startswith(("#", "|", "```", "???", ":::", "<", "- ", "* ", "import "))
|
||||
):
|
||||
continue
|
||||
plain = _plain_markdown(stripped)
|
||||
if plain.startswith(_PLACE_LINES):
|
||||
continue
|
||||
# A paragraph that ends on a colon introduces what follows: alone,
|
||||
# it describes nothing ("One file needs no clone and no local setup:").
|
||||
if len(plain) >= 35 and not plain.endswith(":"):
|
||||
return _shorten(plain)
|
||||
return f"{title}: RetroBIOS source-traced retrogaming firmware reference."
|
||||
if len(plain) >= minimum and not plain.endswith(":"):
|
||||
return plain
|
||||
return ""
|
||||
|
||||
|
||||
# Section headings of a system page that are not a console.
|
||||
_NOT_A_CONSOLE = {"Next steps", "Files", "Other files"}
|
||||
_CARD_TITLE = re.compile(r'class="rb-file-title"[^>]*><code>([^<]+)</code>')
|
||||
|
||||
|
||||
def _system_description(markdown: str, browser: str) -> str:
|
||||
"""A system page by what it holds: a manufacturer page by its consoles,
|
||||
a console page by its "Start with" line, else by its first files."""
|
||||
consoles = [
|
||||
_plain_markdown(re.sub(r"\s*\{[^}]*\}\s*$", "", heading))
|
||||
for heading in re.findall(r"^## (.+?)[ \t]*$", markdown, re.M)
|
||||
]
|
||||
consoles = [name for name in consoles if name and name not in _NOT_A_CONSOLE]
|
||||
if len(consoles) > 1:
|
||||
shown = consoles if len(consoles) <= 4 else consoles[:3] + [f"{len(consoles) - 3} more"]
|
||||
return _shorten(f"{browser} for {_said_as_list(shown)}, one card per file with its "
|
||||
"hashes and the emulators that read it.", 155)
|
||||
start = re.search(r"^\*\*Start with\*\*(?P<line>.*)$(?P<list>(?:\n+- .*)*)", markdown, re.M)
|
||||
if start:
|
||||
# "Start with the files of your machine:" introduces one line per machine.
|
||||
machines = [_plain_markdown(item) for item in re.findall(r"^- (.+)$", start["list"], re.M)]
|
||||
text = _plain_markdown(start["line"]).lstrip(": ")
|
||||
said = "; ".join(machines) if machines and text.endswith(":") else text
|
||||
return _shorten(f"{browser}. Start with {said}", 155)
|
||||
names = list(dict.fromkeys(html.unescape(name) for name in _CARD_TITLE.findall(markdown)))
|
||||
if names:
|
||||
shown = names if len(names) <= 3 else names[:3] + [f"{len(names) - 3} more"]
|
||||
return _shorten(f"{browser}: {_said_as_list(shown)}, with their hashes and the "
|
||||
"emulators that read them.", 155)
|
||||
return _shorten(f"{browser}, one card per file with its hashes and the emulators "
|
||||
"that read it.", 155)
|
||||
|
||||
|
||||
# Structured data a page generator knows beyond its text, by page path
|
||||
@@ -2555,15 +2650,25 @@ LEVELS = {
|
||||
# - A guide behind the first two levels (reference, options, contributing)
|
||||
# names everything a reader types, and ranks after the page that answers.
|
||||
# - A launcher, an alias or a frozen copy answers a name after the emulator.
|
||||
SEARCH_WEIGHT = {"start": 10, "setup": 1.5, "reference": 0.5, "contribute": 0.5}
|
||||
# - A part of a split page (a console's files, 5 of 9) holds cards, each found
|
||||
# by its own heading, and ranks after the page that answers the question.
|
||||
SEARCH_WEIGHT = {"start": 10, "setup": 1.5, "reference": 0.5, "contribute": 0.5, "part": 0.5}
|
||||
SYSTEM_PAGE_WEIGHT = 5
|
||||
DEEP_GUIDE_WEIGHT = 0.5
|
||||
SEARCH_DEMOTED: set[str] = set()
|
||||
# A tool page a reader reaches by a task, not its name: "check my dump",
|
||||
# "md5 check" found emulator pages and never the page that checks a file.
|
||||
SEARCH_WEIGHT_OF_PAGE = {"files/index.md": 2}
|
||||
DEMOTED_WEIGHT = 0.5
|
||||
|
||||
|
||||
def search_weight(relative: Path, level: str) -> float | None:
|
||||
"""The search weight of a page, None for the default."""
|
||||
"""The search weight of a page, None for the default. A part page
|
||||
states its own in its front matter: search_weight(path, "part")."""
|
||||
if level == "part":
|
||||
return SEARCH_WEIGHT["part"]
|
||||
if relative.as_posix() in SEARCH_WEIGHT_OF_PAGE:
|
||||
return SEARCH_WEIGHT_OF_PAGE[relative.as_posix()]
|
||||
if relative.as_posix() in SEARCH_DEMOTED:
|
||||
return DEMOTED_WEIGHT
|
||||
if relative.parts[0] == "systems":
|
||||
@@ -2643,7 +2748,7 @@ def decorate_markdown_pages(docs: Path) -> dict[str, str]:
|
||||
|
||||
title = _page_title(markdown, relative)
|
||||
browser_title = _browser_title(relative, title)
|
||||
description = _page_description(markdown, relative, title)
|
||||
description = str(authored.get("description") or _page_description(markdown, relative, title))
|
||||
if relative.name == "index.md":
|
||||
page_path = "" if len(relative.parts) == 1 else "/".join(relative.parts[:-1]) + "/"
|
||||
else:
|
||||
@@ -2675,6 +2780,7 @@ def decorate_markdown_pages(docs: Path) -> dict[str, str]:
|
||||
search=authored.get("search") != "exclude",
|
||||
logo=str(authored.get("logo") or ""),
|
||||
parent=authored.get("parent"),
|
||||
authored=authored,
|
||||
)
|
||||
write_decorated(docs / target, text)
|
||||
# One file per page: a page that gained or lost its components
|
||||
@@ -8855,7 +8961,7 @@ def generate_nav(
|
||||
],
|
||||
"reference": [
|
||||
# Find a file first: the one tool a reader comes to with a file in hand.
|
||||
(5, {"Find a file": "files/index.md"}, "Checking files"),
|
||||
(5, {"Find or check a file": "files/index.md"}, "Checking files"),
|
||||
(20, {"What the packs do not hold": "gaps.md"}, "Checking files"),
|
||||
(30, {"Cores by platform": "cross-reference.md"}, "Checking files"),
|
||||
(40, {"Dump provenance": "provenance.md"}, "Checking files"),
|
||||
|
||||
Reference in new issue
Block a user