From 946f25035e76bd97c1deb22fd5aec98312d07014 Mon Sep 17 00:00:00 2001
From: ebeshero <ebbondar@gmail.com>
Date: Fri, 28 Apr 2023 01:26:51 -0400
Subject: [PATCH] repairing and restoring XSLT, spot fixes to personTagger.py

---
 pythonCode/personTagger.py      | 14 ++---
 xslt/htmlConverter-complete.xsl | 93 +++++++++++++++++++++++++++++++++
 xslt/xml-paragraph-fixer.xsl    | 17 +-----
 3 files changed, 101 insertions(+), 23 deletions(-)
 create mode 100644 xslt/htmlConverter-complete.xsl

diff --git a/pythonCode/personTagger.py b/pythonCode/personTagger.py
index ff5a25f..e6cf326 100644
--- a/pythonCode/personTagger.py
+++ b/pythonCode/personTagger.py
@@ -1266,11 +1266,11 @@ patterns = [
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "W\.\s+?Citrine"}}]},
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "E\.\s+?Moore"}}]},
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "Lyndon\s+?Johnson"}}]},
-    {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "John\s+?Kennedy"}}]},
+    {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "John(\s*F\.?)\s*Kennedy"}}]},
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "Zia\s+?ul-Haque"}}]},
-    {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "Ronald\s+?Payne"}}]},
-    {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "R\.\s+?Muldoon"}}]},
-    {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "F\.\s+?Orr"}}]},
+    {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "Ronald\s*Payne"}}]},
+    {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "R\.\s*Muldoon"}}]},
+    {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "F\.\s*Orr"}}]},
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "Frank\s+?H\.\s+?Schwable"}}]},
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "R\.\s+?F\.\s+?Doyle"}}]},
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "David\s+?Munson"}}]},
@@ -1382,7 +1382,6 @@ patterns = [
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "Frank\s+?H\.\s+?Schwable"}}]},
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "Billy\s+?Goodman"}}]},
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "Leonard\s+?Pullin"}}]},
-    {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "John\s+?F\.\s+?Kennedy"}}]},
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "R\.\s+?J\.\s+?Biggar"}}]},
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "David\s+?R\.\s+?Hunter"}}]},
     {"label": "PERSON", "pattern": [{"TEXT": {"REGEX": "Hugh\s+?Everett,\s+?III"}}]},
@@ -1523,7 +1522,7 @@ patterns = [
 ruler.add_patterns(patterns)
 
 workingDir = os.getcwd()
-CollPath = os.path.join(workingDir, '../regexConspTest')
+CollPath = os.path.join(workingDir, '../pre-src-xml')
 outputPath = os.path.join(workingDir, 'personTestingOutput/')
 # Everything in original conspiracy directory.
 insideDir = os.listdir(CollPath)
@@ -1636,7 +1635,8 @@ def checkTags(file):
             newLine = regex.sub(r"(<ent type=')<ent type='ORG'>(ORG)</ent>('>)", r"\1\2\3", newLine)
             newLine = regex.sub(r"(<ent type='[A-Z]+'>)<ent type='[A-Z]+'>(\w+)</ent><ent type='[A-Z]+'>(\w+)</ent>(</ent>)", "\1\2 \3\4", newLine)
             newLine = regex.sub(r"(<ent type='[A-Z]+?'>)(\w+)\s+?(<ent type='[A-Z+?]'>)(\w+)(</ent>)(\w+)(<ent type='[A-Z]+?'>)(\w+)(</ent>)(</ent>)", r"\1\2 \4 \6 \8\9", newLine)
-            newLine = regex.sub(r"<ent type='\w+'>(\w+)</ent>('\w)", r"\1\2", newLine)
+            # ebb: Problem line below: eliminates <ent type="PERSON">John Kennedy</ent>'s
+            # newLine = regex.sub(r"<ent type='\w+'>(\w+)</ent>('\w)", r"\1\2", newLine)
 #
 # <spe<ent type='ORG'>cia</ent>l>
 # <<ent type='ORG'>di</ent>v>
diff --git a/xslt/htmlConverter-complete.xsl b/xslt/htmlConverter-complete.xsl
new file mode 100644
index 0000000..f8b60ba
--- /dev/null
+++ b/xslt/htmlConverter-complete.xsl
@@ -0,0 +1,93 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+    xmlns:xs="http://www.w3.org/2001/XMLSchema"
+    xmlns:math="http://www.w3.org/2005/xpath-functions/math"
+    xmlns="http://www.w3.org/1999/xhtml"
+    exclude-result-prefixes="xs math"
+    version="3.0">
+    
+   <xsl:output method="xhtml" html-version="5" omit-xml-declaration="yes" include-content-type="no" indent="yes"/>
+    
+    <!-- 2023-04-27 ebb:  This is XSLT Stage 2: Outputting HTML for the individual collection files -->
+    <xsl:variable name="conspiracy" as="document-node()+" select="collection('../src-xml')"/>
+    
+    
+    <xsl:template match="/">
+        <xsl:for-each select="$conspiracy">   
+            <xsl:variable name="filename" as="xs:string" select="current() ! base-uri() ! tokenize(., '/')[last()] ! substring-before(., '.xml')"/>
+            <xsl:result-document method="xml" indent="yes" href="../docs/collection/{$filename}.html"> 
+    =  <html>
+            <head>
+                <title><xsl:value-of select="$filename"/></title>
+                <link rel="stylesheet" href="../CSSstyle.css"/>
+                <!--Fill in your link line for CSS and JS in the XSLT here! -->
+                <xsl:comment>Fill in your link line for CSS and JS in the XSLT here! </xsl:comment>   
+            </head>
+          <body>
+              <h1 id="title-index"><xsl:value-of select="$filename"/></h1>
+              <nav id="menu">
+                  <a href="../index.html">
+                      <div class="button">Home</div>
+                  </a>
+                  <a href="../fulltext.html">
+                      <div class="button">Fulltext</div>
+                  </a>
+                  <a href="../analysis.html">
+                      <div class="button">Analysis</div>
+                  </a>
+                  <a href="../gallery.html">
+                      <div class="button">Gallery</div>
+                  </a>
+                  <a href="../methods.html">
+                      <div class="button">Methods</div>
+                  </a>
+                  <a href="../about.html">
+                      <div class="button">About</div>
+                  </a>
+                  <a href="../GitHub.html">
+                      <div class="button">GitHub <img alt="github icon"
+                          src="https://logos-download.com/wp-content/uploads/2016/09/GitHub_logo.png"
+                          width="15"/>
+                      </div>
+                  </a>
+              </nav>
+               
+               <xsl:apply-templates/>
+            
+            </body>
+        </html>
+            </xsl:result-document>
+        </xsl:for-each>
+       
+ 
+    </xsl:template>
+    
+    <xsl:template match="p">
+        <p>
+            <xsl:apply-templates/>
+        </p>
+    </xsl:template>
+    <!-- ebb: adding mouseover tooltip via title attribute-->
+    <xsl:template match="ent">
+        <span class="{@type}" title="{@type}">
+            <xsl:apply-templates/>
+        </span>
+    </xsl:template>
+    
+    <!--ebb: What about the special and info XML tags? -->
+    
+    <xsl:template match="special">
+        <span class="special">
+            <xsl:apply-templates/>
+        </span>
+    </xsl:template>
+  <xsl:template match="info">
+      <span class="info" title="{@type}">
+          <xsl:apply-templates/>
+      </span>
+    
+  </xsl:template>
+    
+    
+    
+</xsl:stylesheet>
\ No newline at end of file
diff --git a/xslt/xml-paragraph-fixer.xsl b/xslt/xml-paragraph-fixer.xsl
index 810713e..12b9282 100644
--- a/xslt/xml-paragraph-fixer.xsl
+++ b/xslt/xml-paragraph-fixer.xsl
@@ -21,7 +21,7 @@
 
            <xsl:for-each select="$conspiracy">   
                <xsl:variable name="filename" as="xs:string" select="current() ! base-uri() ! tokenize(., '/')[last()]"/>
-               <xsl:result-document method="xml" indent="yes" href="../src-xml/{$filename}"> 
+               <xsl:result-document method="xml" indent="yes" href="../pre-src-xml/{$filename}"> 
                 <!-- ebb: NEED TO LOOK UP HOW TO SET UP INDIVIDUAL RESULT DOCUMENTS output to folder  -->
                <xsl:choose>
                   <xsl:when test="count(descendant::p) gt 1">
@@ -57,19 +57,4 @@
         </xsl:analyze-string>
         </div>
     </xsl:template>
-    
-    <xsl:template match="info">
-        <info type="{@type}">
-            <xsl:apply-templates/>
-        </info>
-    </xsl:template>
-    <xsl:template match="special">
-        <info type="{@type}">
-            <xsl:apply-templates/>
-        </info>
-    </xsl:template>
-    
-    
-    
-    
 </xsl:stylesheet>
\ No newline at end of file