Run xslt pass 2 and three on the whole year volumes
very fast...
diff --git a/Makefile b/Makefile
index 209b19b..f69482e 100644
--- a/Makefile
+++ b/Makefile
@@ -21,8 +21,11 @@
krill: $(TARGET_DIR)/dnb$(YY).krill.tar
index: $(TARGET_DIR)/dnb$(YY).index.tar.xz
+$(TARGET_DIR)/dnb$(YY).i5.xml: $(TARGET_DIR)/dnb$(YY).pre.i5.xml
+ $(SAXON) -xsl:xslt/pass2.xsl $< | $(SAXON) -xsl:xslt/pass3.xsl - > $@
-$(TARGET_DIR)/dnb$(YY).i5.xml: $(patsubst $(SRC_DIR)/%.epub,$(TARGET_DIR)/%.i5.xml,$(wildcard $(SRC_DIR)/*.epub))
+
+$(TARGET_DIR)/dnb$(YY).pre.i5.xml: $(patsubst $(SRC_DIR)/%.epub,$(TARGET_DIR)/%.i5.xml,$(wildcard $(SRC_DIR)/*.epub))
@echo $(patsubst $(SRC_DIR)/%.epub,$(TARGET_DIR)/%.i5.xml,$(wildcard $(SRC_DIR)/*0.epub)) > $(TARGET_DIR)/filelist$(YY).txt
@echo $(patsubst $(SRC_DIR)/%.epub,$(TARGET_DIR)/%.i5.xml,$(wildcard $(SRC_DIR)/*1.epub)) >> $(TARGET_DIR)/filelist$(YY).txt
@echo $(patsubst $(SRC_DIR)/%.epub,$(TARGET_DIR)/%.i5.xml,$(wildcard $(SRC_DIR)/*2.epub)) >> $(TARGET_DIR)/filelist$(YY).txt
@@ -33,7 +36,7 @@
@echo $(patsubst $(SRC_DIR)/%.epub,$(TARGET_DIR)/%.i5.xml,$(wildcard $(SRC_DIR)/*7.epub)) >> $(TARGET_DIR)/filelist$(YY).txt
@echo $(patsubst $(SRC_DIR)/%.epub,$(TARGET_DIR)/%.i5.xml,$(wildcard $(SRC_DIR)/*8.epub)) >> $(TARGET_DIR)/filelist$(YY).txt
@echo $(patsubst $(SRC_DIR)/%.epub,$(TARGET_DIR)/%.i5.xml,$(wildcard $(SRC_DIR)/*9.epub)) >> $(TARGET_DIR)/filelist$(YY).txt
- sed -i -e 's/ /\n/g' $(TARGET_DIR)/filelist$(YY).txt
+ sed -i -e 's/ /\n/g; /^$$/d' $(TARGET_DIR)/filelist$(YY).txt
head -n -1 xslt/idsCorpus-template.xml | sed -e 's/{YY}/$(YY)/' > $@
@while IFS= read -r f; do \
if head -500 "$$f" | grep -Eq '<pubDate type="year">..$(YY)'; then \
@@ -60,7 +63,7 @@
$(TARGET_DIR)/%.i5.xml: $(BUILD_DIR)/% xslt/epub2i5.xsl xslt/idsCorpus-template.xml
mkdir -p $(TARGET_DIR)
echo "Converting $< to $@"
- $(SAXON) -xsl:xslt/epub2i5.xsl $(shell find $< -name "*.opf") | $(SAXON) -xsl:xslt/unnest_p.xsl - > $@
+ $(SAXON) -xsl:xslt/epub2i5.xsl $(shell find $< -name "*.opf") > $@
%.zip: %.i5.xml
tei2korapxml -l warn -s -tk - < $< > $@
diff --git a/xslt/pass2.xsl b/xslt/pass2.xsl
new file mode 100644
index 0000000..9aef8ca
--- /dev/null
+++ b/xslt/pass2.xsl
@@ -0,0 +1,37 @@
+<xsl:stylesheet version="3.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+ xmlns:saxon="http://saxon.sf.net/"
+ exclude-result-prefixes="saxon">
+
+ <xsl:output method="xml" indent="yes" saxon:line-length="1000"
+ doctype-public="-//IDS//DTD IDS-I5 1.0//EN"
+ doctype-system="http://corpora.ids-mannheim.de/I5/DTD/i5.dtd"
+ />
+
+ <xsl:mode on-no-match="shallow-copy"/>
+
+
+ <xsl:template match="p[descendant::div|descendant::p]" priority="1.0">
+ <div type="section">
+ <xsl:apply-templates/>
+ </div>
+ </xsl:template>
+
+ <xsl:template match="(ref|emph|hi|text())[parent::div]" priority="1.0">
+ <p>
+ <xsl:apply-templates/>
+ </p>
+ </xsl:template>
+
+ <xsl:template match="head[parent::p]">
+ <hi rend="bold">
+ <xsl:value-of select="."/>
+ </hi>
+ </xsl:template>
+
+ <xsl:template match="hi[parent::div]">
+ <p>
+ <xsl:apply-templates/>
+ </p>
+ </xsl:template>
+
+</xsl:stylesheet>
diff --git a/xslt/pass3.xsl b/xslt/pass3.xsl
new file mode 100644
index 0000000..69ae2aa
--- /dev/null
+++ b/xslt/pass3.xsl
@@ -0,0 +1,18 @@
+<xsl:stylesheet version="3.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+ xmlns:saxon="http://saxon.sf.net/"
+ exclude-result-prefixes="saxon">
+
+ <xsl:output method="xml" indent="yes" saxon:line-length="1000"
+ doctype-public="-//IDS//DTD IDS-I5 1.0//EN"
+ doctype-system="http://corpora.ids-mannheim.de/I5/DTD/i5.dtd"
+ />
+
+ <xsl:mode on-no-match="shallow-copy"/>
+
+ <xsl:template match="hi[parent::div]">
+ <p>
+ <xsl:apply-templates/>
+ </p>
+ </xsl:template>
+
+</xsl:stylesheet>
diff --git a/xslt/unnest_p.xsl b/xslt/unnest_p.xsl
deleted file mode 100644
index 26e2cbd..0000000
--- a/xslt/unnest_p.xsl
+++ /dev/null
@@ -1,15 +0,0 @@
-<xsl:stylesheet version="3.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
- xmlns:saxon="http://saxon.sf.net/"
- exclude-result-prefixes="saxon">
-
- <xsl:output method="xml" indent="yes" omit-xml-declaration="yes" saxon:line-length="1000"/>
-
- <xsl:template match="p/p">
- <span>
- <xsl:apply-templates/>
- </span>
- </xsl:template>
-
- <xsl:mode on-no-match="shallow-copy"/>
-
-</xsl:stylesheet>