Fix for escaped JSON string, code improvement and removed verbosity

hotosm · Jan 4, 2024 · 6281af9 · 6281af9
1 parent 103328c
commit 6281af9
Show file tree

Hide file tree

Showing 3 changed files with 48 additions and 19 deletions.
diff --git a/src/data/pq.cc b/src/data/pq.cc
@@ -28,6 +28,7 @@
 #include <iostream>
 #include <string>
 #include <vector>
+#include <sstream>
 
 #include "utils/log.hh"
 using namespace logger;
@@ -170,30 +171,59 @@ std::string Latin1ToUTF8(const std::string& latin1str) {
 }
 
 std::string
-Pq::escapedString(std::string text)
+Pq::escapedString(const std::string &s)
 {
     std::string newstr;
     int i = 0;
-    while (i < text.size()) {
-        if (text[i] == '\'') {
+    while (i < s.size()) {
+        if (s[i] == '\'') {
             newstr += "&apos;";
-        } else if (text[i] == '\"') {
+        } else if (s[i] == '\"') {
             newstr += "&quot;";
-        } else if (text[i] == '\'') {
+        } else if (s[i] == '\'') {
             newstr += "&quot;";
-        } else if (text[i] == ')') {
+        } else if (s[i] == ')') {
             newstr += "&#41;";
-        } else if (text[i] == '(') {
+        } else if (s[i] == '(') {
             newstr += "&#40;";
-        } else if ((text[i] == '\\') || (text[i] == '\n')) {
+        } else if ((s[i] == '\\') || (s[i] == '\n')) {
             // drop this character
         } else {
-            newstr += text[i];
+            newstr += s[i];
         }
         i++;
     }
 
     return sdb->esc(Latin1ToUTF8(newstr));
 }
 
+std::string 
+Pq::escapedJSON(const std::string &s) {
+    std::ostringstream o;
+    for (auto c = s.cbegin(); c != s.cend(); c++) {
+        switch (*c) {
+        case '\x00': o << "\\u0000"; break;
+        case '\x01': o << " "; break;
+        case '\x02': o << " "; break;
+        case '\x03': o << " "; break;
+        case '\x04': o << " "; break;
+        case '\x05': o << " "; break;
+        case '\x06': o << " "; break;
+        case '\x07': o << " "; break;
+        case '\x08': o << " "; break;
+        case '\x09': o << " "; break;
+        case '\x0a': o << "\\n"; break;
+        case '\x0b': o << " "; break;
+        case '\x0c': o << " "; break;
+        case '\x0d': o << " "; break;
+        case '\x0e': o << " "; break;
+        case '\x1f': o << "\\u001f"; break;
+        case '\x22': o << "\\\""; break;
+        case '\x5c': o << "\\\\"; break;
+        default: o << *c;
+        }
+    }
+    return o.str();
+}
+
 } // namespace pq
diff --git a/src/data/pq.hh b/src/data/pq.hh
@@ -64,7 +64,10 @@ class Pq {
     void dump(void);
 
     // Escape string
-    std::string escapedString(std::string text);
+    std::string escapedString(const std::string &s);
+
+    // Escape JSON
+    std::string escapedJSON(const std::string &s);
 
     // Database connection
     std::shared_ptr<pqxx::connection> sdb;

diff --git a/src/raw/queryraw.cc b/src/raw/queryraw.cc
@@ -88,12 +88,13 @@ QueryRaw::applyChange(const OsmNode &node) const
             for (auto it = std::begin(node.tags); it != std::end(node.tags); ++it) {
                 std::string tag_format = "\"%s\" : \"%s\",";
                 boost::format tag_fmt(tag_format);
-                tag_fmt % dbconn->escapedString(it->first);
-                tag_fmt % dbconn->escapedString(it->second);
+                tag_fmt % dbconn->escapedString(dbconn->escapedJSON(it->first));
+                tag_fmt % dbconn->escapedString(dbconn->escapedJSON(it->second));
                 tags += tag_fmt.str();
             }
             tags.erase(tags.size() - 1);
             tags = "'{" + tags + "}'";
+
         } else {
             tags = "null";
         }
@@ -169,8 +170,8 @@ QueryRaw::applyChange(const OsmWay &way) const
                 for (auto it = std::begin(way.tags); it != std::end(way.tags); ++it) {
                     std::string tag_format = "\"%s\" : \"%s\",";
                     boost::format tag_fmt(tag_format);
-                    tag_fmt % dbconn->escapedString(it->first);
-                    tag_fmt % dbconn->escapedString(it->second);
+                    tag_fmt % dbconn->escapedString(dbconn->escapedJSON(it->first));
+                    tag_fmt % dbconn->escapedString(dbconn->escapedJSON(it->second));
                     tags += tag_fmt.str();
                 }
                 tags.erase(tags.size() - 1);
@@ -373,7 +374,6 @@ QueryRaw::getNodeCacheFromWays(std::shared_ptr<std::vector<OsmWay>> ways, std::m
 }
 
 std::map<std::string, std::string> parseTagsString(std::string input) {
-    // std::cout << "[INPUT] " << input << std::endl;
     std::map<std::string, std::string> tags;
     boost::property_tree::ptree pt;
     try {
@@ -386,10 +386,6 @@ std::map<std::string, std::string> parseTagsString(std::string input) {
     for (const auto& pair : pt) {
         tags[pair.first] = pair.second.get_value<std::string>();
     }
-    // for (const auto& pair : tags) {
-    //     std::cout << pair.first << "=" << pair.second << std::endl;
-    // }
-    // std::cout << std::endl;
     return tags;
 }