feature: Add mimetype to knora json response (#305)

* Version set to 2.0.0 * Bugfix: explode dumped core on linux * Added pdf support/mimtype for knora.json
dasch-swiss · Oct 2, 2019 · 81ab98a · 81ab98a
1 parent 9fe6c1e
commit 81ab98a
Show file tree

Hide file tree

Showing 10 changed files with 51 additions and 36 deletions.
diff --git a/CMakeLists.txt b/CMakeLists.txt
@@ -30,7 +30,7 @@ include(CheckIncludeFiles)
 project(sipi)
 
 string(TIMESTAMP SIPIDATE "%Y-%m-%d %H:%M")
-set(SIPI_RELEASE "Sipi Version v1.5.0-SNAPSHOT (build ${SIPIDATE})")
+set(SIPI_RELEASE "Sipi Version v2.0.0-SNAPSHOT (build ${SIPIDATE})")
 
 # ------------------------------------------------------------------------------
 # Set our path variables

diff --git a/config/sipi.init.lua b/config/sipi.init.lua
@@ -308,8 +308,14 @@ function pre_flight(prefix,identifier,cookie)
 
     return 'allow', filepath
 
-end
-    --[[
+--[[
+
+    if config.prefix_as_path then
+        filepath = config.imgroot .. '/' .. prefix .. '/' .. identifier
+    else
+        filepath = config.imgroot .. '/' .. identifier
+    end
+
     if server.cookies['sipi-auth'] then
         print('preflight: IIIF cookie')
         access_info = server.cookies['sipi-auth']
@@ -346,7 +352,6 @@ end
             label = 'Login to SIPI',
         }, filepath
     end
-
-end
 --]]
+end
 -------------------------------------------------------------------------------
diff --git a/include/SipiIO.h b/include/SipiIO.h
@@ -56,8 +56,9 @@ namespace Sipi {
         int width;
         int height;
         int numpages;
+        std::string internalmimetype;
         std::string origname;
-        std::string mimetype;
+        std::string origmimetype;
 
         SipiImgInfo(void) {
             success = FAILURE;

diff --git a/shttps/Global.h b/shttps/Global.h
@@ -40,17 +40,6 @@ namespace shttps {
     }
     //-------------------------------------------------------------------------
 
-    template<class OutIt>
-    void explode(std::string const &input, char sep, OutIt output) {
-        std::istringstream buffer(input);
-        std::string temp;
-
-        while (std::getline(buffer, temp, sep)) {
-            *output++ = temp;
-        }
-    }
-    //-------------------------------------------------------------------------
-
 
     inline std::string getFileName(const std::string &s) {
         char sep = '/';

diff --git a/src/SipiHttpServer.cpp b/src/SipiHttpServer.cpp
@@ -386,7 +386,7 @@ namespace Sipi {
         // test if we have access to the file
         //
         if (access(infile.c_str(), R_OK) != 0) { // test, if file exists
-            throw SipiError(__file__, __LINE__, "Cannot read image file!");
+            throw SipiError(__file__, __LINE__, "Cannot read image file: " + infile);
         }
         pre_flight_info["infile"] = infile;
         return pre_flight_info;
@@ -646,8 +646,9 @@ namespace Sipi {
         if (info.numpages > 0) {
             json_object_set_new(root, "numpages", json_integer(info.numpages));
         }
+        json_object_set_new(root, "internalMimeType", json_string(info.internalmimetype.c_str()));
         if (info.success == SipiImgInfo::ALL) {
-            json_object_set_new(root, "originalMimeType", json_string(info.mimetype.c_str()));
+            json_object_set_new(root, "originalMimeType", json_string(info.origmimetype.c_str()));
             json_object_set_new(root, "originalFilename", json_string(info.origname.c_str()));
         }
         char *json_str = json_dumps(root, JSON_INDENT(3));
@@ -879,7 +880,7 @@ namespace Sipi {
                     (actual_mimetype == "image/jpeg") ||
                     (actual_mimetype == "image/png") ||
                     (actual_mimetype == "image/jpx") ||
-                    (actual_mimetype == "image/jp2")) { // no PDF here!! They are servced as blob
+                    (actual_mimetype == "image/jp2")) { // no PDF here!! They are served as blob
                     conn_obj.setBuffer();
                     conn_obj.status(Connection::SEE_OTHER);
                     const std::string host = conn_obj.header("host");

diff --git a/src/SipiImage.cpp b/src/SipiImage.cpp
@@ -355,17 +355,20 @@ namespace Sipi {
         std::transform(fext.begin(), fext.end(), _fext.begin(), ::tolower);
 
         SipiImgInfo info;
-        if ((_fext == "tif") || (_fext == "tiff")) {
+        std::string mimetype = shttps::Parsing::getFileMimetype(filepath).first;
+
+        if ((mimetype == "image/tiff") || (mimetype == "image/x-tiff")) {
             info = io[std::string("tif")]->getDim(filepath, pagenum);
-        } else if ((_fext == "jpg") || (_fext == "jpeg")) {
+        } else if ((mimetype == "image/jpeg") || (mimetype == "image/pjpeg")) {
             info = io[std::string("jpg")]->getDim(filepath, pagenum);
-        } else if (_fext == "png") {
+        } else if (mimetype == "image/png") {
             info = io[std::string("png")]->getDim(filepath, pagenum);
-        } else if ((_fext == "jp2") || (_fext == "jpx")) {
+        } else if ((mimetype == "image/jp2") || (mimetype == "image/jpx")) {
             info = io[std::string("jpx")]->getDim(filepath, pagenum);
-        } else if (_fext == "pdf") {
+        } else if (mimetype == "application/pdf") {
             info = io[std::string("pdf")]->getDim(filepath, pagenum);
         }
+        info.internalmimetype = mimetype;
 
         if (info.success == SipiImgInfo::FAILURE) {
             for (auto const &iterator : io) {

diff --git a/src/formats/SipiIOJ2k.cpp b/src/formats/SipiIOJ2k.cpp
@@ -648,7 +648,7 @@ namespace Sipi {
             const char *cstr = comment.get_text();
             if (strncmp(cstr, "SIPI:", 5) == 0) {
                 SipiEssentials se(cstr + 5);
-                info.mimetype = se.mimetype();
+                info.origmimetype = se.mimetype();
                 info.origname = se.origname();
                 info.success = SipiImgInfo::ALL;
                 break;

diff --git a/src/formats/SipiIOTiff.cpp b/src/formats/SipiIOTiff.cpp
@@ -1037,7 +1037,7 @@ namespace Sipi {
             char *emdatastr;
             if (1 == TIFFGetField(tif, TIFFTAG_SIPIMETA, &emdatastr)) {
                 SipiEssentials se(emdatastr);
-                info.mimetype = se.mimetype();
+                info.origmimetype = se.mimetype();
                 info.origname = se.origname();
                 info.success = SipiImgInfo::ALL;
             }

diff --git a/src/metadata/SipiEssentials.cpp b/src/metadata/SipiEssentials.cpp
@@ -157,26 +157,41 @@ namespace Sipi {
         _icc_profile = base64Encode(icc_profile_p);
     }
 
+    // for string delimiter
+    std::vector<std::string> split (std::string s, std::string delimiter) {
+        size_t pos_start = 0, pos_end, delim_len = delimiter.length();
+        std::string token;
+        std::vector<std::string> res;
+
+        while ((pos_end = s.find (delimiter, pos_start)) != std::string::npos) {
+            token = s.substr (pos_start, pos_end - pos_start);
+            pos_start = pos_end + delim_len;
+            res.push_back (token);
+        }
+
+        res.push_back (s.substr (pos_start));
+        return res;
+    }
+
     void SipiEssentials::parse(const std::string &str) {
-        std::vector<std::string> result(6);
-        shttps::explode(str, '|', result.begin());
-        _origname = *result.begin();
-        _mimetype = *(result.begin() + 1);
-        std::string _hash_type_str = *(result.begin() + 2);
+        std::vector<std::string> result = split(str, "|");
+        _origname = result[0];
+        _mimetype = result[1];
+        std::string _hash_type_str = result[2];
         if (_hash_type_str == "none") _hash_type = shttps::HashType::none;
         else if (_hash_type_str == "md5") _hash_type = shttps::HashType::md5;
         else if (_hash_type_str == "sha1") _hash_type = shttps::HashType::sha1;
         else if (_hash_type_str == "sha256") _hash_type = shttps::HashType::sha256;
         else if (_hash_type_str == "sha384") _hash_type = shttps::HashType::sha384;
         else if (_hash_type_str == "sha512") _hash_type = shttps::HashType::sha512;
         else _hash_type = shttps::HashType::none;
-        _data_chksum = *(result.begin() + 3);
+        _data_chksum = result[3];
 
         if (result.size() > 5) {
-            std::string tmp_use_icc = *(result.begin() + 4);
+            std::string tmp_use_icc = result[4];
             _use_icc = (tmp_use_icc == "USE_ICC");
             if (_use_icc) {
-                _icc_profile = *(result.begin() + 5);
+                _icc_profile = result[5];
             }
             else {
                 _icc_profile = std::string();

diff --git a/test/e2e/test_02_server.py b/test/e2e/test_02_server.py
@@ -132,7 +132,8 @@ def test_knora_info_validation(self, manager):
             "width": 512,
             "height": 512,
             "originalFilename": "lena512.tif",
-            "originalMimeType": "image/tiff"
+            "originalMimeType": "image/tiff",
+            "internalMimeType": "image/jpx"
         }
 
         response_json = manager.post_file("/api/upload", manager.data_dir_path("unit/lena512.tif"), "image/tiff")