From c51017cf595c6e58c9051348f4eae2cbb5344aa3 Mon Sep 17 00:00:00 2001
From: Daniel Stenberg <daniel@haxx.se>
Date: Thu, 16 May 2024 14:49:42 +0200
Subject: [PATCH] trurl: add "strict:" as prefix to a get component

The strict prefix makes trurl immediately exit with an error code if
such a get can't be done due to URL decoding problems. By default, such
a problem will only make trurl skip that part and silently continue.

Use it like this: --get "{strict:query}"

I also made "url:" a valid prefix, and now we consider the single colon
prefix to be a shortcut for url:. The concept of prefixes scale better
when we use real words instead of single characters.

Fixes #305
Closes #310
---
 tests.json | 28 ++++++++++++++++++++++++++++
 trurl.1    | 13 ++++++++++---
 trurl.c    | 25 +++++++++++++++++--------
 3 files changed, 55 insertions(+), 11 deletions(-)

diff --git a/tests.json b/tests.json
index 844f9085..2e440752 100644
--- a/tests.json
+++ b/tests.json
@@ -2544,5 +2544,33 @@
           "stderr": "",
           "returncode": 0
       }
+  },
+  {
+      "input": {
+          "arguments": [
+              "https://curl.se?name=mr%00smith",
+              "--get",
+              "{query}"
+          ]
+      },
+      "expected": {
+          "stderr": "trurl note: URL decode error, most likely because of rubbish in the input (query)\n",
+          "returncode": 0,
+          "stdout": "\n"
+      }
+  },
+  {
+      "input": {
+          "arguments": [
+              "https://curl.se?name=mr%00smith",
+              "--get",
+              "{strict:query}"
+          ]
+      },
+      "expected": {
+          "stderr": "trurl error: problems URL decoding query\ntrurl error: Try trurl -h for help\n",
+          "returncode": 10,
+          "stdout": ""
+      }
   }
 ]
diff --git a/trurl.1 b/trurl.1
index 2faad488..d3309ccb 100644
--- a/trurl.1
+++ b/trurl.1
@@ -109,10 +109,17 @@ user, password, options, host, port, path, query, fragment and zoneid.
 not have a value.
 
 Components are shown URL decoded by default. If you instead write the
-component prefixed with a colon like "{:path}", it gets output URL encoded.
+component prefixed with "url:" like "{url:path}", it gets output URL encoded.
+As a shortcut, "url:" also works written as a single colon: "{:path}".
 
-You may also prefix components with \fBdefault:\fP and/or \fBpuny:\fP or \fBidn:\fP,
-in any order.
+URL decoding a component may cause problems to display it. Such problems make
+a warning get displayed unless \fB--quiet\fP is used. URL decode problems can
+optionally be turned into errors by prefixing the comopnent name with
+"strict:", as in "{strict:path}". In this stricter mode, a URL decode problem
+makes trurl stop what it is doing and return with exit code 10.
+
+You may also prefix components with \fBdefault:\fP and/or \fBpuny:\fP or
+\fBidn:\fP, in any order.
 
 If \fBdefault:\fP is specified, like "{default:url}" or
 "{default:port}", and the port is not explicitly specified in the URL,
diff --git a/trurl.c b/trurl.c
index 38a49ed0..1ccb3bbf 100644
--- a/trurl.c
+++ b/trurl.c
@@ -757,6 +757,7 @@ static void get(struct option *o, CURLU *uh)
         size_t vlen;
         bool isquery = false;
         bool queryall = false;
+        bool strict = false; /* strict mode, fail on URL decode problems */
         int mods = 0;
         end = strchr(ptr, endbyte);
         ptr++; /* pass the { */
@@ -766,39 +767,44 @@ static void get(struct option *o, CURLU *uh)
           continue;
         }
 
-        /* {path} {:path} */
+        /* {path} {:path} {/path} */
         if(*ptr == ':') {
           mods |= VARMODIFIER_URLENCODED;
           ptr++;
         }
         vlen = end - ptr;
         do {
+          size_t wordlen;
           cl = memchr(ptr, ':', vlen);
           if(!cl)
             break;
+          wordlen = cl - ptr + 1;
 
           /* modifiers! */
-          if(!strncmp(ptr, "default:", cl - ptr + 1))
+          if(!strncmp(ptr, "default:", wordlen))
             mods |= VARMODIFIER_DEFAULT;
-          else if(!strncmp(ptr, "puny:", cl - ptr + 1)) {
+          else if(!strncmp(ptr, "puny:", wordlen)) {
             if(mods & VARMODIFIER_PUNY2IDN)
               errorf(o, ERROR_GET,
                      "puny modifier is mutually exclusive with idn");
             mods |= VARMODIFIER_PUNY;
           }
-          else if(!strncmp(ptr, "idn:", cl - ptr + 1)) {
+          else if(!strncmp(ptr, "idn:", wordlen)) {
             if(mods & VARMODIFIER_PUNY)
               errorf(o, ERROR_GET,
                      "idn modifier is mutually exclusive with puny");
             mods |= VARMODIFIER_PUNY2IDN;
           }
+          else if(!strncmp(ptr, "strict:", wordlen))
+            strict = true;
+          else if(!strncmp(ptr, "url:", wordlen))
+            mods |= VARMODIFIER_URLENCODED;
           else {
-            /* {query: or {query-all: */
-            if(!strncmp(ptr, "query-all:", cl - ptr + 1)) {
+            if(!strncmp(ptr, "query-all:", wordlen)) {
               isquery = true;
               queryall = true;
             }
-            else if(!strncmp(ptr, "query:", cl - ptr + 1))
+            else if(!strncmp(ptr, "query:", wordlen))
               isquery = true;
             else {
               /* syntax error */
@@ -844,7 +850,10 @@ static void get(struct option *o, CURLU *uh)
               /* silently ignore */
               break;
             default:
-              trurl_warnf(o, "%s (%s)", curl_url_strerror(rc), v->name);
+              if((rc == CURLUE_URLDECODE) && strict)
+                errorf(o, ERROR_GET, "problems URL decoding %s", v->name);
+              else
+                trurl_warnf(o, "%s (%s)", curl_url_strerror(rc), v->name);
               break;
             }
           }