Miqra al pi hamasorah importer

README.md

@@ -46,6 +46,13 @@ uv run python -m opensiddur.importer.jps1917.convert_wikisource \
   --project-dir ~/src/opensiddur-repos/opensiddur-projects/project/jps1917
 ```
 
+Example: download Miqra al pi ha-Masorah from Google Sheets into sourcetexts:
+
+```bash
+uv run python -m opensiddur.importer.miqra_al_pi_hamasorah.download \
+  --sourcetexts-root ~/src/opensiddur-repos/sourcetexts/sources
+```
+
 ## JLPTEI sources
 
 JLPTEI sources are compiled into the `project` directory.

codecov.yml

@@ -0,0 +1,11 @@
+# https://docs.codecov.com/docs/commit-status
+coverage:
+  status:
+    project:
+      default:
+        target: 85%
+        informational: false
+    patch:
+      default:
+        target: 85%
+        informational: false

opensiddur/exporter/pdf/pdf.py

@@ -392,5 +392,5 @@ def main():  # pragma: no cover
         sys.exit(1)
 
 
-if __name__ == "__main__":
+if __name__ == "__main__":  # pragma: no cover
     main()

opensiddur/exporter/tex/bibtex.xslt

@@ -3,12 +3,34 @@
     xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
     xmlns:tei="http://www.tei-c.org/ns/1.0"
     xmlns:j="http://jewishliturgy.org/ns/jlptei/2"
+    xmlns:xs="http://www.w3.org/2001/XMLSchema"
     exclude-result-prefixes="tei j">
 
   <xsl:output method="text" encoding="UTF-8" indent="no"/>
 
   <xsl:strip-space elements="*"/>
 
+  <xsl:variable name="xml-ns" select="'http://www.w3.org/XML/1998/namespace'"/>
+
+  <xsl:function name="j:in-scope-lang" as="xs:string">
+    <xsl:param name="node" as="node()"/>
+    <xsl:sequence select="string(($node/ancestor-or-self::*[@xml:lang][1]/@xml:lang)[1])"/>
+  </xsl:function>
+
+  <xsl:function name="j:bibtex-field-value" as="xs:string">
+    <xsl:param name="node" as="element()"/>
+    <xsl:variable name="lang" select="j:in-scope-lang($node)"/>
+    <xsl:variable name="v" select="normalize-space(string($node))"/>
+    <xsl:choose>
+      <xsl:when test="$lang = 'he' or starts-with($lang, 'he-')">
+        <xsl:sequence select="concat('\texthebrew{', $v, '}')"/>
+      </xsl:when>
+      <xsl:otherwise>
+        <xsl:sequence select="$v"/>
+      </xsl:otherwise>
+    </xsl:choose>
+  </xsl:function>
+
   <!-- Root template -->
   <xsl:template match="/">
     <xsl:apply-templates select="//tei:bibl"/>
@@ -215,7 +237,7 @@
       <xsl:text>  </xsl:text>
       <xsl:value-of select="$field-name"/>
       <xsl:text> = {</xsl:text>
-      <xsl:value-of select="normalize-space(.)"/>
+      <xsl:value-of select="j:bibtex-field-value(.)"/>
       <xsl:text>},&#10;</xsl:text>
     </xsl:if>
   </xsl:template>

opensiddur/exporter/tex/latex.py

@@ -464,5 +464,5 @@ def main():  # pragma: no cover
     )
 
 
-if __name__ == "__main__":
+if __name__ == "__main__":  # pragma: no cover
     main()

opensiddur/exporter/tex/reledmac.xslt

@@ -539,7 +539,7 @@
                             <xsl:text>\pend&#10;</xsl:text>
                         </xsl:if>
                         <xsl:text>\eledchapter{</xsl:text>
-                        <xsl:value-of select="f:escape-tex(string(@title))"/>
+                        <xsl:value-of select="f:format-section-title(string(@title), string(@xml:lang))"/>
                         <xsl:text>}&#10;</xsl:text>
                         <xsl:next-iteration>
                             <xsl:with-param name="in-pstart" select="false()"/>
@@ -550,7 +550,7 @@
                             <xsl:text>\pend&#10;</xsl:text>
                         </xsl:if>
                         <xsl:text>\eledsubsection{</xsl:text>
-                        <xsl:value-of select="f:escape-tex(string(@title))"/>
+                        <xsl:value-of select="f:format-section-title(string(@title), string(@xml:lang))"/>
                         <xsl:text>}&#10;</xsl:text>
                         <xsl:next-iteration>
                             <xsl:with-param name="in-pstart" select="false()"/>
@@ -659,17 +659,21 @@
 
     <xsl:template match="tei:body/tei:div" mode="leaves">
         <xsl:if test="tei:head">
+            <xsl:variable name="head" select="tei:head[1]"/>
             <xsl:element name="f:eledpart" namespace="urn:opensiddur:reledmac">
-                <xsl:attribute name="title" select="normalize-space(string-join(tei:head//text(), ''))"/>
+                <xsl:attribute name="title" select="normalize-space(string-join($head//text(), ''))"/>
+                <xsl:attribute name="xml:lang" select="f:section-title-lang($head)"/>
             </xsl:element>
         </xsl:if>
         <xsl:apply-templates select="node()[not(self::tei:head)]" mode="leaves"/>
     </xsl:template>
 
     <xsl:template match="tei:div" mode="leaves" priority="-1">
         <xsl:if test="tei:head">
+            <xsl:variable name="head" select="tei:head[1]"/>
             <xsl:element name="f:eledsubsection" namespace="urn:opensiddur:reledmac">
-                <xsl:attribute name="title" select="normalize-space(string-join(tei:head//text(), ''))"/>
+                <xsl:attribute name="title" select="normalize-space(string-join($head//text(), ''))"/>
+                <xsl:attribute name="xml:lang" select="f:section-title-lang($head)"/>
             </xsl:element>
         </xsl:if>
         <xsl:apply-templates select="node()[not(self::tei:head)]" mode="leaves"/>
@@ -725,7 +729,9 @@
         <!-- Ensure we're in horizontal mode before forcing a linebreak.
              This avoids \"There's no line here to end\" when lb occurs at the
              start of a paragraph/block. -->
-        <xsl:text>\leavevmode\\&#10;</xsl:text>
+        <!-- Add an empty brace group so a following `[` at the start of the next
+             line is not parsed as the optional length argument to `\\`. -->
+        <xsl:text>\leavevmode\\{}&#10;</xsl:text>
     </xsl:template>
 
     <!-- tei:anchor: linkage ids only; editorial notes are already inlined in the body. -->
@@ -893,6 +899,32 @@
         <xsl:sequence select="count($ctx/preceding::tei:note[not(@type='instruction') and not(ancestor::tei:standOff)])"/>
     </xsl:function>
 
+    <!-- Language for tei:head used in \eledchapter/\eledsubsection titles. -->
+    <xsl:function name="f:section-title-lang" as="xs:string">
+        <xsl:param name="head" as="element(tei:head)"/>
+        <xsl:sequence select="string((
+            $head/@xml:lang,
+            $head/ancestor::tei:div[@xml:lang][1]/@xml:lang,
+            $head/ancestor::tei:TEI[@xml:lang][1]/@xml:lang
+        )[1])"/>
+    </xsl:function>
+
+    <!-- Hebrew titles stay in the stream direction; other languages need an
+         explicit LTR wrapper so Latin text is not reversed in RTL blocks. -->
+    <xsl:function name="f:format-section-title" as="xs:string">
+        <xsl:param name="title" as="xs:string"/>
+        <xsl:param name="lang" as="xs:string"/>
+        <xsl:variable name="escaped" select="f:escape-tex($title)"/>
+        <xsl:choose>
+            <xsl:when test="$lang = 'he' or starts-with($lang, 'he-')">
+                <xsl:sequence select="$escaped"/>
+            </xsl:when>
+            <xsl:otherwise>
+                <xsl:sequence select="concat('{\textdir TLT\selectlanguage{english}', $escaped, '}')"/>
+            </xsl:otherwise>
+        </xsl:choose>
+    </xsl:function>
+
     <xsl:function name="f:escape-tex" as="xs:string">
         <xsl:param name="s" as="xs:string"/>
         <!-- Escape characters that have special meaning in LaTeX. Order matters:

opensiddur/exporter/validate_urn_references.py

@@ -178,6 +178,6 @@ def main(argv: Optional[list[str]] = None) -> int:
     return 0
 
 
-if __name__ == "__main__":
+if __name__ == "__main__":  # pragma: no cover
     raise SystemExit(main())
 

opensiddur/importer/agent/checkpoint_example.py

@@ -171,9 +171,10 @@ def demo_interrupted_session():
         print(f"\n🧹 Cleaned up {checkpoint_file}")
 
 
-if __name__ == "__main__":
-    # Run the checkpointing demo
+def _run_cli() -> None:  # pragma: no cover
     demo_checkpointing()
-
-    # Run the interrupted session demo
     demo_interrupted_session()
+
+
+if __name__ == "__main__":  # pragma: no cover
+    _run_cli()

opensiddur/importer/agent/example_usage.py

@@ -63,5 +63,5 @@ def main():
             print(f"  [{role}]: {message[:100]}{'...' if len(message) > 100 else ''}")
 
 
-if __name__ == "__main__":
+if __name__ == "__main__":  # pragma: no cover
     main()