Index pdf annotations separately under field name annotation. Add annot, pdfannot and pa aliases.

2020-10-12 10:05:38 +02:00
parent f60851e935
commit 25eda37bc9
4 changed files with 35 additions and 1 deletions
@@ -443,13 +443,17 @@ class PDFExtractor:
        #self.em.rclog("Annotations: %s" % abypage)
        pagevec = html.split(b"\f")
        html = b""
+        annotsfield = b""
        pagenum = 1
        for page in pagevec:
            html += page
            if pagenum in abypage:
                html += abypage[pagenum].encode('utf-8')
+                annotsfield += abypage[pagenum].encode('utf-8') + b" - " 
            html += b"\f"
            pagenum += 1
+        if annotsfield:
+            self.em.setfield("pdfannot", annotsfield)
        return html
    
    def _selfdoc(self):