notes_mdda.txt

 Host hs-github
    User git
    Hostname github.com
    IdentityFile ~/.ssh/id_rsa@github-hs

[remote "origin"]
#url = git@github.com:HandshakesByDC/poppler.git
 url = hs-github:HandshakesByDC/poppler.git
fetch = +refs/heads/*:refs/remotes/origin/*

dnf install openjpeg2-devel

git clone git@github.com:HandshakesByDC/poppler.git
cd poppler
mkdir -p build
cd build
rm -f CMakeCache.txt
cmake ..
make


#GOODPDF=~/SEER/services/textextract/server/cache/DA350094-CB0B-4C05-9444-762313D4AB98/0/0-original.pdf
#NULLPDF=~/SEER/services/pdftodoc/0-original_1571026944371705_CrashOnWilby.pdf

GOODPDF=../test-pdfs/Ace_Achieve_AR2014.pdf
NULLPDF=../test-pdfs/0-original_1571026944371705_NullFontNames.pdf

# Bad Fails on 0.37+mdda = N/a
# Both Work on git head > 0.83.0 = 39baa7d42966ebd67c2ac91ef1c1450965c37e87
TESTPDF=${GOODPDF}
TESTPDF=${NULLPDF}
./utils/pdftotext -q -r 72 -bbox -enc UTF-8 ${TESTPDF} -

# TESTS
./utils/pdftotext -q -r 72 -bbox -enc UTF-8 ${GOODPDF} - | grep '<word xMin="79.34" xMax="116.23" yMin="489.09" yMax="496.53">forty-eight</word>'
## Expect :
#no word list (page 1 empty)
#no word list (page 2 empty)
#no word list (page 89 empty)
#  <word xMin="79.34" xMax="116.23" yMin="489.09" yMax="496.53">forty-eight</word>

./utils/pdftotext -q -r 72 -bbox -enc UTF-8 ${NULLPDF} - | grep '<word xMin="90.00" xMax="105.57" yMin="576.55" yMax="585.88">Tel:</word>'
## Expect :
#  <word xMin="90.00" xMax="105.57" yMin="576.55" yMax="585.88">Tel:</word>



----- Add this one - NOOP-ed out ----- 

[andrewsm@square poppler-mdda]$ git diff
diff --git a/poppler/Gfx.cc b/poppler/Gfx.cc
index 07d95b35..0b39bdf5 100644
--- a/poppler/Gfx.cc
+++ b/poppler/Gfx.cc
@@ -3937,6 +3937,7 @@ void Gfx::doShowText(GooString *s) {
   wMode = font->getWMode();
 
   if (out->useDrawChar()) {
+    //printf("ADASDAD"); // mdda  YES: This is used by HtmlOutputDev
     out->beginString(state, s);
   }
 
----- Ignore following file changes ----- 

@@ -4036,6 +4037,8 @@ void Gfx::doShowText(GooString *s) {
 
   } else if (out->useDrawChar()) {
     p = s->getCString();
+    //printf("-%s", p);   // mdda  (A Clue!) : I-NTRO-DUC-T-I-ON    -T-H-E- C-O-D-E-
+      
     len = s->getLength();
     while (len > 0) {
       n = font->getNextChar(p, len, &code,
@@ -4062,6 +4065,24 @@ void Gfx::doShowText(GooString *s) {
       if (ocState)
         out->drawChar(state, state->getCurX() + riseX, state->getCurY() + riseY,
                      tdx, tdy, tOriginX, tOriginY, code, n, u, uLen);

----- Add this one - NOOP-ed out ----- 

+
+// mdda : This is from TextOutputDev.cc#4574
+//     out->drawChar(state, sel->word->textMat[i].m[4], sel->word->textMat[i].m[5], 0, 0, 0, 0,
+//                   sel->word->charcode[i], 1, NULL, 0);
+
+  if(0) { // mdda START     -1I+1I-1n+1N-1t+1T-1r+1R-1o+1O-1D+1D-1U+1U-1C+1C-1t+1T-1I+1I-1o+1O-1n+1N
+          if(1) { // mdda START   (-I-n-t-r-o-D-U-C-t-I-o-n)
+            printf("-%d", uLen);
+            for (int i = 0; i < uLen; ++i) {
+              printf("%c", u[i]);
+            }
+          } // mdda END
+          if(1) { // mdda START   (+I+N+T+R+O+D+U+C+T+I+O+N) 
+            printf("+%d%c", n, code);
+          } // mdda END
+  } // mdda END
+                      
+                      
       state->shift(tdx, tdy);
       p += n;
       len -= n;
@@ -4099,6 +4120,8 @@ void Gfx::doShowText(GooString *s) {
     state->textTransformDelta(dx, dy, &tdx, &tdy);
     if (ocState)
       out->drawString(state, s);
+
+    //printf("BOZO"); // mdda (not called for HTMLOutputDev
     state->shift(tdx, tdy);
   }
 
----- Ignore following file changes ----- 

diff --git a/poppler/GfxFont.cc b/poppler/GfxFont.cc
index 81f59037..347e661b 100644
--- a/poppler/GfxFont.cc
+++ b/poppler/GfxFont.cc
@@ -1666,6 +1666,9 @@ int Gfx8BitFont::getNextChar(char *s, int len, CharCode *code,
   *uLen = ctu->mapToUnicode(c, u);
   *dx = widths[c];
   *dy = *ox = *oy = 0;
+  
+//  printf(" %c", *code);   // mdda 
+  
   return 1;
 }
 
@@ -2241,6 +2244,8 @@ int GfxCIDFont::getNextChar(char *s, int len, CharCode *code,
   *ox = vx;
   *oy = vy;
 
+  // printf("$%c", *code);   // mdda == Only odd characters
+
   return n;
 }
 
----- Ignore following file changes ----- 

diff --git a/utils/HtmlOutputDev.cc b/utils/HtmlOutputDev.cc
index 35e98b46..793073b7 100644
--- a/utils/HtmlOutputDev.cc
+++ b/utils/HtmlOutputDev.cc
@@ -235,6 +235,11 @@ HtmlString::~HtmlString() {
 
 void HtmlString::addChar(GfxState *state, double x, double y,
                         double dx, double dy, Unicode u) {
+                                 
+  if(0) { // mdda START  
+    printf("%c", u);  // HtmlString::addChar  (IntroDUCtIon)
+  } // mdda END
+                                 
   if (dir == textDirUnknown) {
     //dir = UnicodeMap::getDirection(u);
     dir = textDirLeftRight;
@@ -370,6 +375,12 @@ void HtmlPage::addChar(GfxState *state, double x, double y,
   int n, i;
   state->transform(x, y, &x1, &y1);
   n = curStr->len;
+
+  if(0) { // mdda START   (IntroDUCtIon)
+    for (i = 0; i < uLen; ++i) {
+      printf("%c", u[i]);
+    }
+  } // mdda END
  
   // check that new character is in the same direction as current string
   // and is not too far away from it before adding 
@@ -410,6 +421,7 @@ void HtmlPage::addChar(GfxState *state, double x, double y,
   }
   for (i = 0; i < uLen; ++i) {
     curStr->addChar(state, x1 + i*w1, y1 + i*h1, w1, h1, u[i]);
+    //printf("%c", u[i]); // mdda addChar  (IntroDUCtIon)
   }
 }
 
 
----- Don't add this one - Not used by pdftotext ----- 

@@ -763,10 +775,46 @@ void HtmlPage::dumpAsXML(FILE* f,int page){
 
   for(HtmlString *tmp=yxStrings;tmp;tmp=tmp->yxNext){
     if (tmp->htext){
+      // mdda new START  :: Remove trailing space, since it can screw up column widths
+      char *cstr = tmp->htext->getCString();
+      double width = (tmp->xMax-tmp->xMin);
+      int len = strlen(cstr);
+      if(1){ // mdda
+        if(tmp->len>1 && len>1 && cstr[len-1] == ' ') {
+          width -= (tmp->xRight[tmp->len-1] - tmp->xRight[tmp->len-2]);
+       }
+      }
+      // mdda new END
+      
+      // mdda new START  :: Check for 'l o n d o n  S c h o o l' spacing problem
+      if(1){ // mdda
+        if(len>3 && cstr[1] == ' ') {  // Second character a space is quickest check
+          int all_spaced=true;
+          for(int j=3; j<len; j+=2) {
+            if(cstr[j] != ' ') {
+              all_spaced=false;
+            }
+          }
+          if(all_spaced) {
+            //printf("\nAll Spaced : %s\n", cstr);
+            for(int j=0; j<len; j+=2) {
+              cstr[j/2] = cstr[j];
+              cstr[len/2] = '\0';  // terminate it
+            }
+            //printf("Re- Spaced : %s\n", cstr);
+          }
+        }
+      }
+      // mdda new END
+      
       fprintf(f,"<text top=\"%d\" left=\"%d\" ",xoutRound(tmp->yMin),xoutRound(tmp->xMin));
-      fprintf(f,"width=\"%d\" height=\"%d\" ",xoutRound(tmp->xMax-tmp->xMin),xoutRound(tmp->yMax-tmp->yMin));
+      
+      //fprintf(f,"width=\"%d\" height=\"%d\" ",xoutRound(tmp->xMax-tmp->xMin),xoutRound(tmp->yMax-tmp->yMin)); // mdda DEL
+      fprintf(f,"width=\"%d\" height=\"%d\" ",xoutRound(width),xoutRound(tmp->yMax-tmp->yMin));  // mdda NEW
+      
       fprintf(f,"font=\"%d\">", tmp->fontpos);
-      fputs(tmp->htext->getCString(),f);
+      fputs(cstr,f);  // mdda NEW
+      //fputs(tmp->htext->getCString(),f);  // mdda DEL
       fputs("</text>\n",f);
     }
   }

----- Don't add this one - Not used by pdftotext ----- 

@@ -1312,12 +1360,38 @@ void HtmlOutputDev::endString(GfxState *state) {
 void HtmlOutputDev::drawChar(GfxState *state, double x, double y,
              double dx, double dy,
              double originX, double originY,
-             CharCode code, int /*nBytes*/, Unicode *u, int uLen) 
+             //CharCode code, int /*nBytes*/, Unicode *u, int uLen)  // mdda DEL
+             CharCode code, int nBytes, Unicode *u, int uLen)  // mdda NEW
 {
+  Unicode u_code = 0; // mdda 
+
   if ( !showHidden && (state->getRender() & 3) == 3) {
     return;
   }
-  pages->addChar(state, x, y, dx, dy, originX, originY, u, uLen);
+  
+  if(0) { // mdda START   (IntroDUCtIon)
+    for (int i = 0; i < uLen; ++i) {
+      printf("%c", u[i]);  // drawChar
+    }
+  } // mdda END
+  if(0) { // mdda START   (=I=N=T=R=O=D=U=C=T=I=O=N)
+    printf("=%c", code);  // drawChar
+  } // mdda END
+
+  if(1) {// mdda bodge START  -- This checks to see whether the unicode is just length=1, and then substitutes in the 'code' instead
+    if(nBytes == 1 && ((code & 0x7f) == code)) {
+      u_code = code;
+      pages->addChar(state, x, y, dx, dy, originX, originY, &u_code, 1);
+    } 
+    else {
+      pages->addChar(state, x, y, dx, dy, originX, originY, u, uLen);
+    }
+  }
+  else {
+    pages->addChar(state, x, y, dx, dy, originX, originY, u, uLen);   
+  }  // mdda bodge END
+  
+  //pages->addChar(state, x, y, dx, dy, originX, originY, u, uLen);   // mdda DEL
 }
 
diff --git a/utils/pdftotext.cc b/utils/pdftotext.cc
index d931a969..579a83f7 100644
--- a/utils/pdftotext.cc
+++ b/utils/pdftotext.cc
@@ -37,6 +37,7 @@
 #include <stdlib.h>
 #include <stddef.h>
 #include <string.h>
+#include <math.h>  // mdda
 #include "parseargs.h"
 #include "printencodings.h"
 #include "goo/GooString.h"
@@ -516,7 +517,7 @@ void printLine(FILE *f, TextLine *line) {

 
----- Don't add this one - Not used by HS pdftotext ----- 

 void printDocBBox(FILE *f, PDFDoc *doc, TextOutputDev *textOut, int first, int last) {
   double xMin, yMin, xMax, yMax;
-  TextPage *textPage;
+  //TextPage *textPage;
   TextFlow *flow;
   TextBlock *blk;
   TextLine *line;
@@ -525,7 +526,7 @@ void printDocBBox(FILE *f, PDFDoc *doc, TextOutputDev *textOut, int first, int l
   for (int page = first; page <= last; ++page) {
     fprintf(f, "  <page width=\"%f\" height=\"%f\">\n",doc->getPageMediaWidth(page), doc->getPageMediaHeight(page));
     doc->displayPage(textOut, page, resolution, resolution, 0, gTrue, gFalse, gFalse);
-    textPage = textOut->takeText();
+    TextPage *textPage = textOut->takeText();
     for (flow = textPage->getFlows(); flow; flow = flow->getNext()) {
       fprintf(f, "    <flow>\n");
       for (blk = flow->getBlocks(); blk; blk = blk->getNext()) {
@@ -544,6 +545,8 @@ void printDocBBox(FILE *f, PDFDoc *doc, TextOutputDev *textOut, int first, int l
   fprintf(f, "</doc>\n");
 }
 
+#define MDDA_VERSION 1
+#ifndef MDDA_VERSION
 void printWordBBox(FILE *f, PDFDoc *doc, TextOutputDev *textOut, int first, int last) {
   fprintf(f, "<doc>\n");
   for (int page = first; page <= last; ++page) {
@@ -567,3 +570,101 @@ void printWordBBox(FILE *f, PDFDoc *doc, TextOutputDev *textOut, int first, int
   }
   fprintf(f, "</doc>\n");
 }
+#endif
+
+#ifdef MDDA_VERSION
+void printWordBBox(FILE *f, PDFDoc *doc, TextOutputDev *textOut, int first, int last) {
+  fprintf(f, "<doc>\n");
+  
+  GooString *invalid_font_name = new GooString("-INVALID-FONT-NAME-");
+
+  for (int page = first; page <= last; ++page) {
+    fprintf(f, " <page n=\"%d\" width=\"%.2f\" height=\"%.2f\">\n", page, doc->getPageMediaWidth(page), doc->getPageMediaHeight(page));
+    doc->displayPage(textOut, page, resolution, resolution, 0, gTrue, gFalse, gFalse);
+    TextWordList *wordlist = textOut->makeWordList();
+    const int word_length = wordlist != NULL ? wordlist->getLength() : 0;
+    
+    // All the _prev values must be nonsense, to ensure that they are reset (and emitted) at the beginning of each page
+    
+    TextWord *word;
+    double xMinA, xMaxA, yMinA, yMaxA, yMinA_prev=-1.0, yMaxA_prev=-1.0;
+    
+    TextFontInfo *fontinfo;
+    int is_bold, is_italic, is_bold_prev=-1, is_italic_prev=-1; 
+    double fontsize, fontsize_prev=-1.0;
+    double cr,cg,cb, cr_prev=-1,cg_prev=-1,cb_prev=-1;
+    
+    GBool space_after;
+    GooString *fontname, *fontname_prev = invalid_font_name;
+    
+    if (word_length == 0)
+      fprintf(stderr, "no word list (page %d empty)\n", page);
+
+    for (int i = 0; i < word_length; ++i) {
+      std::stringstream wordXML;
+      wordXML << std::fixed << std::setprecision(2);
+
+      std::stringstream wordRGB;
+      wordRGB << std::hex << std::setfill('0') << std::setw(2) << std::right;
+      
+      word = wordlist->get(i);
+      word->getBBox(&xMinA, &yMinA, &xMaxA, &yMaxA);
+      const std::string myString = myXmlTokenReplace(word->getText()->getCString());
+
+      fontinfo = word->getFontInfo(0);
+      is_bold = fontinfo->isBold() ? 1:0;
+      is_italic = fontinfo->isItalic() ?1:0;
+
+      fontsize = word->getFontSize(); 
+      fontname = word->getFontName(0);
+      word->getColor(&cr,&cg,&cb);
+    
+      space_after = word->hasSpaceAfter();
+      
+      wordXML << "  <word xMin=\"" << xMinA << "\" xMax=\"" << xMaxA << "\"";
+      
+      if( fabs(yMinA-yMinA_prev)>0.05 || fabs(yMaxA-yMaxA_prev)>0.05 ) {
+        wordXML << " yMin=\"" << yMinA << "\" yMax=\"" << yMaxA << "\"";
+        yMinA_prev = yMinA;
+        yMaxA_prev = yMaxA;
+      }
+      
+      if( is_bold != is_bold_prev ) {
+        wordXML << " b=\"" << (is_bold?"1":"0") << "\"";
+        is_bold_prev = is_bold;
+      }
+      
+      if( is_italic != is_italic_prev ) {
+        wordXML << " i=\"" << (is_italic?"1":"0") << "\"";
+        is_italic_prev = is_italic;
+      }
+      
+      if( fontsize != fontsize_prev || fontname->cmp(fontname_prev)!=0 ) {
+        wordXML << " fontSize=\"" << fontsize << "\" fontName=\"" << fontname->getCString() << "\"";
+        fontsize_prev = fontsize;
+        fontname_prev = fontname;
+      }
+      
+      if( cr!=cr_prev || cg!=cg_prev || cb!=cb_prev ) {
+        wordRGB << (int)(cr*255) << (int)(cg*255) << (int)(cb*255);
+        wordXML << " rgb=\"" << wordRGB.str() << "\"";
+        cr_prev=cr;
+        cg_prev=cg;
+        cb_prev=cb;
+        //wordRGB.str(""); // Clear the stream?
+      }
+      
+      if( !space_after ) {
+        wordXML << " noSpace=\"1\"";
+      }
+      
+      wordXML << ">" << myString << "</word>\n";
+      
+      fputs(wordXML.str().c_str(), f);
+    }
+    fprintf(f, " </page>\n");
+    delete wordlist;
+  }
+  fprintf(f, "</doc>\n");
+}
+#endif
[andrewsm@square poppler-mdda]$