From 01d0f4b73411f2701b943af48edd0d92f60a888d Mon Sep 17 00:00:00 2001
From: Jean-Francois Dockes <jfd@recoll.org>
Date: Fri, 12 Aug 2016 18:00:52 +0200
Subject: [PATCH] execm filters: the change to let filters set arbitrary
 metadata lost the top doc size, now saved aside

---
 src/common/cstr.h             |  1 -
 src/internfile/internfile.cpp | 38 +++++++++++++++++------------------
 src/internfile/mimehandler.h  |  9 ++++++---
 3 files changed, 24 insertions(+), 24 deletions(-)

diff --git a/src/common/cstr.h b/src/common/cstr.h
index c3c55ca1..7f41099b 100644
--- a/src/common/cstr.h
+++ b/src/common/cstr.h
@@ -80,7 +80,6 @@ DEF_CSTR(dj_keycontent, "content");
 DEF_CSTR(dj_keyipath, "ipath");
 DEF_CSTR(dj_keymd5, "md5");
 DEF_CSTR(dj_keymt, "mimetype");
-DEF_CSTR(dj_keydocsize, "docsize");
 DEF_CSTR(dj_keyanc, "rclanc");
 
 #endif /* _CSTR_H_INCLUDED_ */
diff --git a/src/internfile/internfile.cpp b/src/internfile/internfile.cpp
index 8037adf5..37cf8c5a 100644
--- a/src/internfile/internfile.cpp
+++ b/src/internfile/internfile.cpp
@@ -517,9 +517,9 @@ bool FileInterner::dijontorcl(Rcl::Doc& doc)
 		// point if the last container filter is directly
 		// returning text/plain content, so that there is no
 		// ipath-less filter at the top
-		char cbuf[30];
-		sprintf(cbuf, "%d", int(doc.text.length()));
-		doc.fbytes = cbuf;
+                lltodecstr(doc.text.length(), doc.fbytes);
+                LOGDEB("FileInterner::dijontorcl: fbytes->" << doc.fbytes <<
+                       endl);
 	    }
 	} else if (it->first == cstr_dj_keymd) {
 	    doc.dmtime = it->second;
@@ -584,23 +584,21 @@ void FileInterner::collectIpathAndMT(Rcl::Doc& doc) const
     for (vector<RecollFilter*>::const_iterator hit = m_handlers.begin();
 	 hit != m_handlers.end(); hit++) {
 	const map<string, string>& docdata = (*hit)->get_meta_data();
-	if (getKeyValue(docdata, cstr_dj_keyipath, ipathel)) {
-	    if (!ipathel.empty()) {
-		// Non-empty ipath. This stack element is for an
-		// actual embedded document, not a format translation.
-		hasipath = true;
-		getKeyValue(docdata, cstr_dj_keymt, doc.mimetype);
-		getKeyValue(docdata, cstr_dj_keyfn, doc.meta[Rcl::Doc::keyfn]);
-	    } else {
-		if (doc.fbytes.empty())
-		    getKeyValue(docdata, cstr_dj_keydocsize, doc.fbytes);
-	    }
-	    doc.ipath += colon_hide(ipathel) + cstr_isep;
-	} else {
-	    if (doc.fbytes.empty())
-		getKeyValue(docdata, cstr_dj_keydocsize, doc.fbytes);
-	    doc.ipath += cstr_isep;
-	}
+        ipathel.clear();
+	getKeyValue(docdata, cstr_dj_keyipath, ipathel);
+        if (!ipathel.empty()) {
+            // Non-empty ipath. This stack element is for an
+            // actual embedded document, not a format translation.
+            hasipath = true;
+            getKeyValue(docdata, cstr_dj_keymt, doc.mimetype);
+            getKeyValue(docdata, cstr_dj_keyfn, doc.meta[Rcl::Doc::keyfn]);
+        } else {
+            if (doc.fbytes.empty()) {
+                lltodecstr((*hit)->get_docsize(), doc.fbytes);
+                LOGDEB("collectIpath..: fbytes->" << doc.fbytes << endl);
+            }
+        }
+        doc.ipath += colon_hide(ipathel) + cstr_isep;
         // We set the author field from the innermost doc which has
         // one: allows finding, e.g. an image attachment having no
         // metadata by a search on the sender name. Only do this for
diff --git a/src/internfile/mimehandler.h b/src/internfile/mimehandler.h
index 97257a17..fa03cecf 100644
--- a/src/internfile/mimehandler.h
+++ b/src/internfile/mimehandler.h
@@ -92,9 +92,11 @@ public:
 	return set_document_string(mtype, std::string(cp, sz));
     }
 
-    virtual void set_docsize(off_t size)
-    {
-	m_metaData[cstr_dj_keydocsize] = ulltodecstr(size);
+    virtual void set_docsize(off_t size) {
+	m_docsize = size;
+    }
+    virtual off_t get_docsize() const {
+	return m_docsize;
     }
 
     virtual bool has_documents() const {return m_havedoc;}
@@ -146,6 +148,7 @@ protected:
     // m_id is and md5 of the filter definition line (from mimeconf) and
     // is used when fetching/returning filters to / from the cache.
     std::string m_id;
+    off_t m_docsize; // Size of the top document
 };
 
 /**