Add options to the "diff" command to ignore certain attributes

When the new command line options --ignore-changeset, --ignore-uid, or
--ignore-user are used, the respective attributes of OSM objects are
ignored when comparing the input files. This is especially useful for
user names, because OSM users can change their own user names, which
happens often enough to matter.

When used with the OPL output format the ignored attributes are not
written out to the output.

Author: joto

man/osmium-diff.md

@@ -63,6 +63,19 @@ None of the output formats print the headers of the input files.
 -f, \--output-format=FORMAT
 :   See the **OUTPUT FORMATS** section.
 
+\--ignore-changeset
+:   Ignore changeset id on OSM objects when comparing files. When used with
+    the OPL output format the 'changeset' attribute is not written to the
+    output.
+
+\--ignore-uid
+:   Ignore user id on OSM objects when comparing files. When used with
+    the OPL output format the 'uid' attribute is not written to the output.
+
+\--ignore-user
+:   Ignore user name on OSM objects when comparing files. When used with
+    the OPL output format the 'user' attribute is not written to the output.
+
 -o, \--output=FILE
 :   Name of the output file. Default is '-' (STDOUT).
 
@@ -107,7 +120,7 @@ None of the output formats print the headers of the input files.
 
 # EXAMPLES
 
-Show difference between Nepal files from January 2016 and Febrary 2016 in
+Show difference between Nepal files from January 2016 and February 2016 in
 compact format:
 
     osmium diff nepal-20160101.osm.pbf nepal-20160201.osm.pbf

src/command_diff.cpp

@@ -52,6 +52,9 @@ along with this program.  If not, see <https://www.gnu.org/licenses/>.
 bool CommandDiff::setup(const std::vector<std::string>& arguments) {
     po::options_description opts_cmd{"COMMAND OPTIONS"};
     opts_cmd.add_options()
+    ("ignore-changeset", "Ignore changeset id when comparing objects")
+    ("ignore-uid", "Ignore user id when comparing objects")
+    ("ignore-user", "Ignore user name when comparing objects")
     ("object-type,t", po::value<std::vector<std::string>>(), "Read only objects of given type (node, way, relation)")
     ("output,o", po::value<std::string>(), "Output file")
     ("output-format,f", po::value<std::string>(), "Format of output file")
@@ -93,6 +96,18 @@ bool CommandDiff::setup(const std::vector<std::string>& arguments) {
         throw argument_error("You need exactly two input files for this command.");
     }
 
+    if (vm.count("ignore-changeset")) {
+        m_ignore_attrs_changeset = true;
+    }
+
+    if (vm.count("ignore-uid")) {
+        m_ignore_attrs_uid = true;
+    }
+
+    if (vm.count("ignore-user")) {
+        m_ignore_attrs_user = true;
+    }
+
     if (vm.count("output")) {
         m_output_filename = vm["output"].as<std::string>();
     }
@@ -130,6 +145,18 @@ bool CommandDiff::setup(const std::vector<std::string>& arguments) {
             m_output_file = osmium::io::File{m_output_filename, m_output_format};
             m_output_file.check();
 
+            std::string metadata{"version+timestamp"};
+            if (!m_ignore_attrs_changeset) {
+                metadata += "+changeset";
+            }
+            if (!m_ignore_attrs_uid) {
+                metadata += "+uid";
+            }
+            if (!m_ignore_attrs_changeset) {
+                metadata += "+user";
+            }
+            m_output_file.set("add_metadata", metadata);
+
             auto f = m_output_file.format();
             if (f != osmium::io::file_format::opl && f != osmium::io::file_format::debug) {
                 throw argument_error("File format does not support diff output. Use 'compact', 'opl' or 'debug' format.");
@@ -248,6 +275,21 @@ class OutputActionOSM : public OutputAction {
 
 }; // class OutputActionOSM
 
+void CommandDiff::update_object_crc(osmium::CRC<osmium::CRC_zlib>* crc, const osmium::OSMObject &object) {
+    crc->update_bool(object.visible());
+    crc->update(object.timestamp());
+    crc->update(object.tags());
+    if (!m_ignore_attrs_changeset) {
+        crc->update_int32(object.changeset());
+    }
+    if (!m_ignore_attrs_uid) {
+        crc->update_int32(object.uid());
+    }
+    if (!m_ignore_attrs_user) {
+        crc->update_string(object.user());
+    }
+}
+
 bool CommandDiff::run() {
     osmium::io::Reader reader1{m_input_files[0], osm_entity_bits()};
     osmium::io::ReaderWithProgressBar reader2{display_progress(), m_input_files[1], osm_entity_bits()};
@@ -298,18 +340,20 @@ bool CommandDiff::run() {
         } else { /* *it1 == *it2 */
             osmium::CRC<osmium::CRC_zlib> crc1;
             osmium::CRC<osmium::CRC_zlib> crc2;
+            update_object_crc(&crc1, *it1);
+            update_object_crc(&crc2, *it2);
             switch (it1->type()) {
                 case osmium::item_type::node:
-                    crc1.update(static_cast<const osmium::Node&>(*it1));
-                    crc2.update(static_cast<const osmium::Node&>(*it2));
+                    crc1.update(static_cast<const osmium::Node&>(*it1).location());
+                    crc2.update(static_cast<const osmium::Node&>(*it2).location());
                     break;
                 case osmium::item_type::way:
-                    crc1.update(static_cast<const osmium::Way&>(*it1));
-                    crc2.update(static_cast<const osmium::Way&>(*it2));
+                    crc1.update(static_cast<const osmium::Way&>(*it1).nodes());
+                    crc2.update(static_cast<const osmium::Way&>(*it2).nodes());
                     break;
                 case osmium::item_type::relation:
-                    crc1.update(static_cast<const osmium::Relation&>(*it1));
-                    crc2.update(static_cast<const osmium::Relation&>(*it2));
+                    crc1.update(static_cast<const osmium::Relation&>(*it1).members());
+                    crc2.update(static_cast<const osmium::Relation&>(*it2).members());
                     break;
                 default:
                     break;

src/command_diff.hpp

@@ -25,15 +25,23 @@ along with this program.  If not, see <https://www.gnu.org/licenses/>.
 
 #include "cmd.hpp" // IWYU pragma: export
 
+#include <osmium/osm/crc.hpp>
+#include <osmium/osm/crc_zlib.hpp>
+
 #include <string>
 #include <vector>
 
 class CommandDiff : public CommandWithMultipleOSMInputs, public with_osm_output {
 
     std::string m_output_action;
+    bool m_ignore_attrs_changeset = false;
+    bool m_ignore_attrs_uid = false;
+    bool m_ignore_attrs_user = false;
     bool m_show_summary = false;
     bool m_suppress_common = false;
 
+    void update_object_crc(osmium::CRC<osmium::CRC_zlib>* crc, const osmium::OSMObject &object);
+
 public:
 
     explicit CommandDiff(const CommandFactory& command_factory) :

test/diff/CMakeLists.txt

@@ -20,4 +20,12 @@ check_diff(opl-c "-f opl -c" input1.osm input2.osm output-c.opl 1)
 
 check_diff(same "" input1.osm input1.osm output-same 0)
 
+check_diff(ignore-uid "--ignore-uid -c" input1.osm input2.osm output-compact-c-nouid 1)
+
+check_diff(ignore-uid-15 "--ignore-uid -c" input1uid.osm input2uid.osm output-empty 0)
+
+check_diff(ignore-uid-opl "--ignore-uid -f opl -c" input1.osm input2.osm output-compact-c-nouid-opl 1)
+
+check_diff(ignore-uid-15-opl "--ignore-uid -f opl -c" input1uid.osm input2uid.osm output-empty 0)
+
 #-----------------------------------------------------------------------------

test/diff/input1uid.osm

@@ -0,0 +1,4 @@
+<?xml version='1.0' encoding='UTF-8'?>
+<osm version="0.6" upload="false" generator="testdata">
+  <node id="15" version="1" timestamp="2015-01-01T02:00:00Z" uid="2" user="test" changeset="1" lat="5" lon="1"/>
+</osm>

test/diff/input2.osm

@@ -1,5 +1,5 @@
 <?xml version='1.0' encoding='UTF-8'?>
-<osm version="0.6" generator="test">
+<osm version="0.6" upload="false" generator="testdata">
   <node id="10" version="1" timestamp="2015-01-01T01:00:00Z" uid="1" user="test" changeset="1" lat="1" lon="1"/>
   <node id="11" version="2" timestamp="2015-01-01T02:00:00Z" uid="1" user="test" changeset="2" lat="2" lon="2"/>
   <node id="12" version="1" timestamp="2015-01-01T01:00:00Z" uid="1" user="test" changeset="1" lat="3" lon="1"/>

test/diff/input2uid.osm

@@ -0,0 +1,4 @@
+<?xml version='1.0' encoding='UTF-8'?>
+<osm version="0.6" upload="false" generator="testdata">
+  <node id="15" version="1" timestamp="2015-01-01T02:00:00Z" uid="1" user="test" changeset="1" lat="5" lon="1"/>
+</osm>

test/diff/output-compact-c-nouid

@@ -0,0 +1,7 @@
+-n11 v1
++n11 v2
+-n13 v1
++n14 v1
+*n16 v1
+-w21 v1
++w21 v2

test/diff/output-compact-c-nouid-opl

@@ -0,0 +1,8 @@
+-n11 v1 dV c1 t2015-01-01T01:00:00Z utest T x1 y2
++n11 v2 dV c2 t2015-01-01T02:00:00Z utest T x2 y2
+-n13 v1 dV c1 t2015-01-01T01:00:00Z utest T x1 y4
++n14 v1 dV c2 t2015-01-01T02:00:00Z utest T x1 y5
+-n16 v1 dV c1 t2015-01-01T02:00:00Z utest T x2 y5
++n16 v1 dV c1 t2015-01-01T02:00:00Z utest T x1 y5
+-w21 v1 dV c1 t2015-01-01T01:00:00Z utest Txyz=abc Nn12,n13
++w21 v2 dV c2 t2015-01-01T02:00:00Z utest Txyz=new Nn12,n14