Add s-104 and t-076 for possible <dfn> candidates in endnotes

apasel422 · Jul 13, 2024 · cc86fb0 · cc86fb0
1 parent 4ea1530
commit cc86fb0
Show file tree

Hide file tree

Showing 6 changed files with 99 additions and 0 deletions.
diff --git a/se/se_epub_lint.py b/se/se_epub_lint.py
@@ -352,6 +352,7 @@
 "s-101", "Anonymous primary contributor value not exactly [text]Anonymous[/]."
 "s-102", "[attr]lang[/] attribute detected. Hint: Use [attr]xml:lang[/] instead."
 "s-103", "Probable missing semantics for a roman I numeral."
+"s-104", "[xhtml]<i>[/] element without semantic at beginning of endnote. Hint: Should this be [xhtml]<dfn>[/] instead?"
 
 TYPOGRAPHY
 "t-001", "Double spacing found. Sentences should be single-spaced. (Note that double spaces might include Unicode no-break spaces!)"
@@ -429,6 +430,7 @@
 "t-073", "Possible transcription error in Greek."
 "t-074", "Extended sound using hyphen-minus [text]-[/] instead of non-breaking hyphen [text]‑[/]."
 "t-075", "Word in verse with acute accent for scansion instead of grave accent."
+"t-076", "[xhtml]<dfn>[/] at beginning of endnote should be set in italics."
 
 
 XHTML
@@ -1500,6 +1502,15 @@ def _lint_special_file_checks(self, filename: Path, dom: se.easy_xml.EasyXmlTree
 		if nodes:
 			messages.append(LintMessage("t-027", "Endnote backlink not preceded by exactly one space.", se.MESSAGE_TYPE_WARNING, filename, [node.to_string() for node in nodes]))
 
+		# Check for semantic-less <i> at beginning of endnotes
+		nodes = dom.xpath("/html/body//li[contains(@epub:type, 'endnote') and ./p/node()[1][name() = 'i' and not(@epub:type) and string-length(.) < 20 and string-length(./following-sibling::text()) > 20]]")
+		if nodes:
+			messages.append(LintMessage("s-104", "[xhtml]<i>[/] element without semantic at beginning of endnote. Hint: Should this be [xhtml]<dfn>[/] instead?", se.MESSAGE_TYPE_ERROR, filename, [node.to_string() for node in nodes]))
+
+		# <dfn> in endnotes without font-style: italic
+		if dom.xpath("/html/body//li[contains(@epub:type, 'endnote') and ./p/node()[1][name() = 'dfn' and not(@data-css-font-style = 'italic')]]"):
+			messages.append(LintMessage("t-076", "[xhtml]<dfn>[/] at beginning of endnote should be set in italics.", se.MESSAGE_TYPE_WARNING, filename))
+
 	# Check LoI descriptions to see if they match associated figcaptions
 	elif special_file == "loi":
 		for node in dom.xpath("/html/body/nav[contains(@epub:type, 'loi')]//li//a"):

diff --git a/tests/lint/semantic/s-104/golden/s-104-out.txt b/tests/lint/semantic/s-104/golden/s-104-out.txt
@@ -0,0 +1,19 @@
+m-033 [Manual Review] endnotes.xhtml `endnotes` semantic inflection found, but 
+no MARC relator `ann` (Annotator).
+s-056 [Error] endnotes.xhtml Last `<p>` child of endnote missing backlink.
+        <li epub:type="endnote">
+        <li epub:type="endnote">
+        <li epub:type="endnote">
+        <li epub:type="endnote">
+        <li epub:type="endnote">
+        <li epub:type="endnote">
+s-104 [Error] endnotes.xhtml `<i>` element without semantic at beginning of 
+endnote. Hint: Should this be `<dfn>` instead?
+        <li epub:type="endnote">
+                <p><i>Cho</i> is generally translated as “street.”</p>
+        </li>
+        <li epub:type="endnote">
+                <p>Some other text.</p>
+                <p><i>Tsu</i> is an instance of shortening a name in familiar 
+speech.</p>
+        </li>
diff --git a/tests/lint/semantic/s-104/in/src/epub/text/endnotes.xhtml b/tests/lint/semantic/s-104/in/src/epub/text/endnotes.xhtml
@@ -0,0 +1,39 @@
+<?xml version="1.0" encoding="utf-8"?>
+<html xmlns="http://www.w3.org/1999/xhtml" xmlns:epub="http://www.idpf.org/2007/ops" epub:prefix="z3998: http://www.daisy.org/z3998/2012/vocab/structure/, se: https://standardebooks.org/vocab/1.0" xml:lang="en-GB">
+	<head>
+		<title>Endnotes</title>
+		<link href="../css/core.css" rel="stylesheet" type="text/css"/>
+		<link href="../css/local.css" rel="stylesheet" type="text/css"/>
+	</head>
+	<body epub:type="backmatter">
+		<section id="endnotes" epub:type="endnotes">
+			<h2 epub:type="title">Endnotes</h2>
+			<ol>
+				<li epub:type="endnote">
+					<p><i>Cho</i> is generally translated as “street.”</p>
+				</li>
+				<!-- s-104 looks at each <p> within an endnote -->
+				<li epub:type="endnote">
+					<p>Some other text.</p>
+					<p><i>Tsu</i> is an instance of shortening a name in familiar speech.</p>
+				</li>
+				<!-- not relevant for s-104 because the <i> has an epub:type -->
+				<li epub:type="endnote">
+					<p><i epub:type="se:name.vessel.ship">Nautilus</i> is a submarine.</p>
+				</li>
+				<!-- not relevant for s-104 because the <i> text is too long -->
+				<li epub:type="endnote">
+					<p><i>012345678901234567891</i> is 21 characters long.</p>
+				</li>
+				<!-- not relevant for s-104 because the following text is too short -->
+				<li epub:type="endnote">
+					<p><i>X</i> 12345678901234567.</p>
+				</li>
+				<!-- not relevant for s-104 because the <i> is not the first child of the <p> -->
+				<li epub:type="endnote">
+					<p>An <i>X</i> marked the spot.</p>
+				</li>
+			</ol>
+		</section>
+	</body>
+</html>
diff --git a/tests/lint/typography/t-076/golden/t-076-out.txt b/tests/lint/typography/t-076/golden/t-076-out.txt
@@ -0,0 +1,6 @@
+m-033 [Manual Review] endnotes.xhtml `endnotes` semantic inflection found, but 
+no MARC relator `ann` (Annotator).
+s-056 [Error] endnotes.xhtml Last `<p>` child of endnote missing backlink.
+        <li epub:type="endnote">
+t-076 [Manual Review] endnotes.xhtml `<dfn>` at beginning of endnote should be 
+set in italics.
diff --git a/tests/lint/typography/t-076/in/src/epub/css/local.css b/tests/lint/typography/t-076/in/src/epub/css/local.css
@@ -0,0 +1,6 @@
+@charset "utf-8";
+
+/* browser.css defaults this to italic, so override it here */
+dfn{
+	font-style: normal;
+}
diff --git a/tests/lint/typography/t-076/in/src/epub/text/endnotes.xhtml b/tests/lint/typography/t-076/in/src/epub/text/endnotes.xhtml
@@ -0,0 +1,18 @@
+<?xml version="1.0" encoding="utf-8"?>
+<html xmlns="http://www.w3.org/1999/xhtml" xmlns:epub="http://www.idpf.org/2007/ops" epub:prefix="z3998: http://www.daisy.org/z3998/2012/vocab/structure/, se: https://standardebooks.org/vocab/1.0" xml:lang="en-GB">
+	<head>
+		<title>Endnotes</title>
+		<link href="../css/core.css" rel="stylesheet" type="text/css"/>
+		<link href="../css/local.css" rel="stylesheet" type="text/css"/>
+	</head>
+	<body epub:type="backmatter">
+		<section id="endnotes" epub:type="endnotes">
+			<h2 epub:type="title">Endnotes</h2>
+			<ol>
+				<li epub:type="endnote">
+					<p><dfn>Cho</dfn> is generally translated as “street.”</p>
+				</li>
+			</ol>
+		</section>
+	</body>
+</html>