Add textunknownasplain configuration variable to tell the indexer to process text/xxx where xxx is unknown as text/plain

src/common/rclconfig.cpp

@@ -1,4 +1,4 @@
-/* Copyright (C) 2004 J.F.Dockes 
+/* Copyright (C) 2004-2022 J.F.Dockes 
  *   This program is free software; you can redistribute it and/or modify
  *   it under the terms of the GNU General Public License as published by
  *   the Free Software Foundation; either version 2 of the License, or
@@ -33,14 +33,11 @@
 #endif
 
 #include <algorithm>
-#include <list>
 #include <iostream>
 #include <sstream>
-#include <fstream>
 #include <cstdlib>
 #include <cstring>
 #include <unordered_map>
-#include <iterator>
 
 #include "cstr.h"
 #include "pathut.h"
@@ -854,6 +851,13 @@ string RclConfig::getMimeHandlerDef(const string &mtype, bool filtertypes, const
     }
 
     if (!mimeconf->get(mtype, hs, "index")) {
+        if (mtype.find("text/") == 0) {
+            bool alltext{false};
+            getConfParam("textunknownasplain", &alltext);
+            if (alltext && mimeconf->get("text/plain", hs, "index")) {
+                return hs;
+            }
+        }
         if (mtype != "inode/directory") {
             IdxDiags::theDiags().record(IdxDiags::NoHandler, fn, mtype);
             LOGDEB1("getMimeHandlerDef: no handler for '" << mtype << "' (fn " << fn << ")\n");
@@ -1233,6 +1237,16 @@ string RclConfig::getMimeViewerDef(const string &mtype, const string& apptag, bo
 
     if (apptag.empty() || !mimeview->get(mtype + string("|") + apptag, hs, "view"))
         mimeview->get(mtype, hs, "view");
+
+    // Last try for text/xxx if alltext is set
+    if (hs.empty() && mtype.find("text/") == 0 && mtype != "text/plain") {
+        bool alltext{false};
+        getConfParam("textunknownasplain", &alltext);
+        if (alltext) {
+            return getMimeViewerDef("text/plain", apptag, useall);
+        }
+    }
+        
     return hs;
 }
 
@@ -1425,7 +1439,7 @@ string RclConfig::getPidfile() const
 #endif // ! _WIN32
     
         fn = path_cat(getCacheDir(), "index.pid");
-out:
+    out:
         LOGINF("RclConfig: pid/lock file: " << fn << "\n");
     }
     return fn;

src/internfile/mimehandler.cpp

@@ -178,11 +178,12 @@ static RecollFilter *mhFactory(RclConfig *config, const string &mimeOrParams,
         MD5String("MimeHandlerNull", id);
         return nobuild ? 0 : new MimeHandlerNull(config, id);
     } else if (lmime.find("text/") == 0) {
-        // Try to handle unknown text/xx as text/plain. This
-        // only happen if the text/xx was defined as "internal" in
-        // mimeconf, not at random. For programs, for example this
-        // allows indexing and previewing as text/plain (no filter
-        // exec) but still opening with a specific editor.
+        // Try to handle unknown text/xx as text/plain. This only happen if the text/xx was defined
+        // as "internal" in mimeconf.
+        // This is historic, because the natural way with the current code would be to specify
+        // "internal text/plain" instead of just "internal"...
+        // Also see rclconfig for the textunknownisplain configuration variable which allows
+        // processing all unknown text/* as text/plain.
         LOGDEB2("mhFactory(" << lmime << "): returning MimeHandlerText(x)\n");
         MD5String("MimeHandlerText", id);
         return nobuild ? 0 : new MimeHandlerText(config, id);

src/sampleconf/mimemap

@@ -39,6 +39,7 @@
 .bas = text/plain 
 .cpp = text/x-c
 .h   = text/x-c
+.hpp = text/x-c++
 .c   = text/x-c
 .cc  = text/x-c
 .cxx = text/x-c

src/sampleconf/recoll.conf

@@ -15,8 +15,7 @@
 # quoted with "double quotes". Double quotes processing is NOT applied to single value parameter,
 # and using quotes there will result in bad values.
 
-# <grouptitle id="WHATDOCS">Parameters affecting what documents we
-# index</grouptitle> 
+# <grouptitle id="WHATDOCS">Parameters affecting what documents we index</grouptitle> 
 
 # <var name="topdirs" type="string"><brief>Space-separated list of files or
 # directories to recursively index.</brief><descr>Default to ~ (indexes
@@ -227,6 +226,15 @@ compressedfilemaxkbs = 100000
 # disable.</descr></var>
 textfilemaxmbs = 20
 
+# <var name="textunknownasplain" type="bool">
+# <brief>Process unknown text/xxx files as text/plain</brief>
+# <descr>Allows indexing misc. text files identified as text/whatever by 'file' or 'xdg-mime'
+# without having to explicitely set config entries for them. This works fine for indexing (but will
+# cause processing of a lot of garbage though), but the documents indexed this way will be opened by
+# the desktop viewer, even if text/plain has a specific editor.</descr>
+# </var>
+#textunknownasplain=0
+
 # <var name="indexallfilenames" type="bool"><brief>Index the file names of
 # unprocessed files</brief><descr>Index the names of files the contents of
 # which we don't index because of an excluded or unsupported MIME