v1.0.42: конвертация .doc через ВМ + парсер .doc

parser.cfm

@@ -163,6 +163,28 @@
     <cfreturn out>
 </cffunction>
 
+<!--- ═══════════════════════════════════════════════════════════
+     Helper: parseDOC — грубый экстрактор текста из .doc
+     ═══════════════════════════════════════════════════════════ --->
+<cffunction name="parseDOC" access="private" returntype="struct">
+    <cfargument name="fileBytes" type="binary">
+    <cfset var out = {elements: [], errors: []}>
+    <cftry>
+        <cfset var raw = createObject("java", "java.lang.String").init(fileBytes, "windows-1251")>
+        <cfset var lines = raw.split("\n")>
+        <cfloop array="#lines#" index="var line">
+            <cfset line = trim(line)>
+            <cfif len(line) GT 3 AND (reFind("[а-яА-ЯёЁ]", line) OR reFind("\d{2,}", line))>
+                <cfset arrayAppend(out.elements, {type: "paragraph", style: "", text: line, page: 0})>
+            </cfif>
+        </cfloop>
+        <cfcatch>
+            <cfset arrayAppend(out.errors, ".doc: " & cfcatch.message)>
+        </cfcatch>
+    </cftry>
+    <cfreturn out>
+</cffunction>
+
 <!--- ═══════════════════════════════════════════════ MAIN --->
 <cfset result = {ok: false, error: ""}>
 <cftry>
@@ -210,7 +232,7 @@
                         <cfset parsed = {elements: [], errors: []}>
                         <cfif zext EQ "pdf">
                             <cfset parsed = parsePDF(entryBytes)>
-                        <cfelseif zext EQ "docx">
+                        <cfelseif zext EQ "docx" OR zext EQ "doc">
                             <cfset parsed = parseDOCX(entryBytes)>
                         <cfelse>
                             <cfset arrayAppend(parsed.errors, "unsupported: .#zext#")>
@@ -264,7 +286,10 @@
                 </cftry>
 
             <cfelseif doc.mime_type EQ "application/msword">
-                <cfset arrayAppend(errors, ".doc not supported (use libreoffice)")>
+                <cfset fileBytes = binaryDecode(doc.original_bytes, "base64")>
+                <cfset parsed = parseDOC(fileBytes)>
+                <cfset elements = parsed.elements>
+                <cfset errors = parsed.errors>
             <cfelse>
                 <cfset arrayAppend(errors, "unsupported: " & doc.mime_type)>
             </cfif>