fix: Fix uri encoding/decoding and add new tests

This currently seems to compile exactly the same in 1.12.0 and 1.13.0 so
therefore has a performance regression

.github/workflows/test.yml

@@ -14,7 +14,7 @@ jobs:
       - uses: actions/checkout@v4
       - uses: erlef/setup-beam@v1
         with:
-          otp-version: "28.0.4"
+          otp-version: "28.1"
           gleam-version: "1.12.0"
           rebar3-version: "3.25.1"
           elixir-version: "1.18.4"

.gitignore

@@ -2,3 +2,4 @@
 *.ez
 /build
 erl_crash.dump
+node_modules

CHANGELOG.md

@@ -12,3 +12,17 @@
 
 - Improved parsing performance significantly and reduced memory usage up to 50%
 - Significantly improved IPV4 parsing performance
+
+## v2.0.2
+
+- Minor performance improvement for uris with userinfo
+- More performance improvements for ascii/digit parsing
+
+## 2.0.3
+
+- Minor performance improvement for erlang
+- Major performance improvement for js
+
+## 2.0.4
+
+- Reverted some optimisations as they are unnecessary for Gleam v1.14.0+

gleam.toml

@@ -1,12 +1,12 @@
 name = "gluri"
-version = "2.0.1"
+version = "2.0.4"
 
 # Fill out these fields if you intend to generate HTML documentation or publish
 # your project to the Hex package manager.
 #
  description = "Uri (RFC 3986) library for Gleam"
  licences = ["Apache-2.0"]
- repository = { type = "github", user = "pendletong", repo = "uri" }
+ repository = { type = "gitea", host = "git.pendleton.ie", user = "pendletong", repo = "uri" }
  links = [{ title = "RFC 3986", href = "https://www.ietf.org/rfc/rfc3986.txt" }]
 #
 # For a full reference of all the available options, you can have a look at

manifest.toml

@@ -3,7 +3,7 @@
 
 packages = [
   { name = "argv", version = "1.0.2", build_tools = ["gleam"], requirements = [], otp_app = "argv", source = "hex", outer_checksum = "BA1FF0929525DEBA1CE67256E5ADF77A7CDDFE729E3E3F57A5BDCAA031DED09D" },
-  { name = "benchee", version = "1.4.0", build_tools = ["mix"], requirements = ["deep_merge", "statistex", "table"], otp_app = "benchee", source = "hex", outer_checksum = "299CD10DD8CE51C9EA3DDB74BB150F93D25E968F93E4C1FA31698A8E4FA5D715" },
+  { name = "benchee", version = "1.5.0", build_tools = ["mix"], requirements = ["deep_merge", "statistex", "table"], otp_app = "benchee", source = "hex", outer_checksum = "5B075393AEA81B8AE74EADD1C28B1D87E8A63696C649D8293DB7C4DF3EB67535" },
   { name = "bigben", version = "1.0.1", build_tools = ["gleam"], requirements = ["birl", "gleam_erlang", "gleam_otp", "gleam_stdlib"], otp_app = "bigben", source = "hex", outer_checksum = "190E489610A80D76C48BACC75EB8314BD184FF0220AB0F251ABE760B993B91BB" },
   { name = "birl", version = "1.8.0", build_tools = ["gleam"], requirements = ["gleam_regexp", "gleam_stdlib", "ranger"], otp_app = "birl", source = "hex", outer_checksum = "2AC7BA26F998E3DFADDB657148BD5DDFE966958AD4D6D6957DD0D22E5B56C400" },
   { name = "deep_merge", version = "1.0.0", build_tools = ["mix"], requirements = [], otp_app = "deep_merge", source = "hex", outer_checksum = "CE708E5F094B9CD4E8F2BE4F00D2F4250C4095BE93F8CD6D018C753894885430" },
@@ -14,9 +14,9 @@ packages = [
   { name = "gleam_erlang", version = "1.3.0", build_tools = ["gleam"], requirements = ["gleam_stdlib"], otp_app = "gleam_erlang", source = "hex", outer_checksum = "1124AD3AA21143E5AF0FC5CF3D9529F6DB8CA03E43A55711B60B6B7B3874375C" },
   { name = "gleam_javascript", version = "1.0.0", build_tools = ["gleam"], requirements = ["gleam_stdlib"], otp_app = "gleam_javascript", source = "hex", outer_checksum = "EF6C77A506F026C6FB37941889477CD5E4234FCD4337FF0E9384E297CB8F97EB" },
   { name = "gleam_json", version = "3.0.2", build_tools = ["gleam"], requirements = ["gleam_stdlib"], otp_app = "gleam_json", source = "hex", outer_checksum = "874FA3C3BB6E22DD2BB111966BD40B3759E9094E05257899A7C08F5DE77EC049" },
-  { name = "gleam_otp", version = "1.1.0", build_tools = ["gleam"], requirements = ["gleam_erlang", "gleam_stdlib"], otp_app = "gleam_otp", source = "hex", outer_checksum = "7987CBEBC8060B88F14575DEF546253F3116EBE2A5DA6FD82F38243FCE97C54B" },
+  { name = "gleam_otp", version = "1.2.0", build_tools = ["gleam"], requirements = ["gleam_erlang", "gleam_stdlib"], otp_app = "gleam_otp", source = "hex", outer_checksum = "BA6A294E295E428EC1562DC1C11EA7530DCB981E8359134BEABC8493B7B2258E" },
   { name = "gleam_regexp", version = "1.1.1", build_tools = ["gleam"], requirements = ["gleam_stdlib"], otp_app = "gleam_regexp", source = "hex", outer_checksum = "9C215C6CA84A5B35BB934A9B61A9A306EC743153BE2B0425A0D032E477B062A9" },
-  { name = "gleam_stdlib", version = "0.63.1", build_tools = ["gleam"], requirements = [], otp_app = "gleam_stdlib", source = "hex", outer_checksum = "E1D5EC07638F606E48F0EA1556044DD805F2ACE9092A6F6AFBE4A0CC4DA21C2F" },
+  { name = "gleam_stdlib", version = "0.65.0", build_tools = ["gleam"], requirements = [], otp_app = "gleam_stdlib", source = "hex", outer_checksum = "7C69C71D8C493AE11A5184828A77110EB05A7786EBF8B25B36A72F879C3EE107" },
   { name = "gleam_time", version = "1.4.0", build_tools = ["gleam"], requirements = ["gleam_stdlib"], otp_app = "gleam_time", source = "hex", outer_checksum = "DCDDC040CE97DA3D2A925CDBBA08D8A78681139745754A83998641C8A3F6587E" },
   { name = "gleam_yielder", version = "1.1.0", build_tools = ["gleam"], requirements = ["gleam_stdlib"], otp_app = "gleam_yielder", source = "hex", outer_checksum = "8E4E4ECFA7982859F430C57F549200C7749823C106759F4A19A78AEA6687717A" },
   { name = "gleeunit", version = "1.6.1", build_tools = ["gleam"], requirements = ["gleam_stdlib"], otp_app = "gleeunit", source = "hex", outer_checksum = "FDC68A8C492B1E9B429249062CD9BAC9B5538C6FBF584817205D0998C42E1DAC" },

package-lock.json

@@ -0,0 +1,18 @@
+{
+  "name": "uri",
+  "lockfileVersion": 3,
+  "requires": true,
+  "packages": {
+    "": {
+      "dependencies": {
+        "mitata": "^1.0.34"
+      }
+    },
+    "node_modules/mitata": {
+      "version": "1.0.34",
+      "resolved": "https://registry.npmjs.org/mitata/-/mitata-1.0.34.tgz",
+      "integrity": "sha512-Mc3zrtNBKIMeHSCQ0XqRLo1vbdIx1wvFV9c8NJAiyho6AjNfMY8bVhbS12bwciUdd1t4rj8099CH3N3NFahaUA==",
+      "license": "MIT"
+    }
+  }
+}

package.json

@@ -0,0 +1,6 @@
+{
+  "type": "module",
+  "dependencies": {
+    "mitata": "^1.0.34"
+  }
+}

src/gluri/internal/parser.gleam

@@ -1,4 +1,3 @@
-import gleam/bool
 import gleam/int
 import gleam/list
 import gleam/option.{type Option, None, Some}
@@ -168,7 +167,7 @@ fn parse_authority(str: String) -> Result(#(Uri, String), Nil) {
 }
 
 fn parse_authority_part(str: String) -> Result(#(Uri, String), Nil) {
-  let #(userinfo, rest) = parse_userinfo(str, "")
+  let #(userinfo, rest) = parse_userinfo(str)
 
   use #(host, rest) <- result.try(parse_host(rest))
 
@@ -180,8 +179,14 @@ fn parse_authority_part(str: String) -> Result(#(Uri, String), Nil) {
 }
 
 // userinfo      = *( unreserved / pct-encoded / sub-delims / ":" )
-fn parse_userinfo(str: String, userinfo: String) -> #(Option(String), String) {
+fn parse_userinfo(str: String) -> #(Option(String), String) {
-  use <- bool.guard(when: !string.contains(str, "@"), return: #(None, str))
+  case string.contains(str, "@") {
+    True -> do_parse_userinfo(str, "")
+    False -> #(None, str)
+  }
+}
+
+fn do_parse_userinfo(str: String, userinfo: String) -> #(Option(String), String) {
   case str {
     "@" <> rest -> #(Some(userinfo), rest)
     "" -> #(None, userinfo <> str)
@@ -202,7 +207,7 @@ fn parse_userinfo(str: String, userinfo: String) -> #(Option(String), String) {
           str,
         )
       {
-        Ok(#(part, rest)) -> parse_userinfo(rest, userinfo <> part)
+        Ok(#(part, rest)) -> do_parse_userinfo(rest, userinfo <> part)
         Error(_) -> #(None, userinfo <> str)
       }
     }
@@ -421,7 +426,7 @@ fn parse_ipv4address(str: String) {
 //              / "1" 2DIGIT            ; 100-199
 //              / "2" %x30-34 DIGIT     ; 200-249
 //              / "25" %x30-35          ; 250-255
-pub fn parse_dec_octet(str: String) -> Result(#(String, String), Nil) {
+fn parse_dec_octet(str: String) -> Result(#(String, String), Nil) {
   try_parsers(
     [
       parse_this_then(_, [
@@ -486,7 +491,7 @@ pub fn parse_dec_octet(str: String) -> Result(#(String, String), Nil) {
 }
 
 // reg-name      = *( unreserved / pct-encoded / sub-delims )
-pub fn parse_reg_name(str: String) {
+fn parse_reg_name(str: String) {
   // can't error
 
   case do_parse_reg_name(str, "") {
@@ -722,19 +727,21 @@ fn parse_unreserved(str: String) -> Result(#(String, String), Nil) {
 
 // sub-delims    = "!" / "$" / "&" / "'" / "(" / ")"
 //               / "*" / "+" / "," / ";" / "="
-fn parse_sub_delim(str: String) {
+//                 %21 / %24 / %26 / %27 / %28 / %29
+//               / %2A / %2B / %2C / %3B / %3D
+fn parse_sub_delim(str: String) -> Result(#(String, String), Nil) {
   case str {
-    "!" as l <> rest
+    "!" as char <> tail
-    | "$" as l <> rest
+    | "$" as char <> tail
-    | "&" as l <> rest
+    | "&" as char <> tail
-    | "'" as l <> rest
+    | "'" as char <> tail
-    | "(" as l <> rest
+    | "(" as char <> tail
-    | ")" as l <> rest
+    | ")" as char <> tail
-    | "*" as l <> rest
+    | "*" as char <> tail
-    | "+" as l <> rest
+    | "+" as char <> tail
-    | "," as l <> rest
+    | "," as char <> tail
-    | ";" as l <> rest
+    | ";" as char <> tail
-    | "=" as l <> rest -> Ok(#(l, rest))
+    | "=" as char <> tail -> Ok(#(char, tail))
     _ -> Error(Nil)
   }
 }
@@ -742,31 +749,34 @@ fn parse_sub_delim(str: String) {
 // DIGIT    = %x30–39
 fn parse_digit(str: String) -> Result(#(String, String), Nil) {
   case str {
-    "0" as l <> rest
+    "0" as char <> tail
-    | "1" as l <> rest
+    | "1" as char <> tail
-    | "2" as l <> rest
+    | "2" as char <> tail
-    | "3" as l <> rest
+    | "3" as char <> tail
-    | "4" as l <> rest
+    | "4" as char <> tail
-    | "5" as l <> rest
+    | "5" as char <> tail
-    | "6" as l <> rest
+    | "6" as char <> tail
-    | "7" as l <> rest
+    | "7" as char <> tail
-    | "8" as l <> rest
+    | "8" as char <> tail
-    | "9" as l <> rest -> Ok(#(l, rest))
+    | "9" as char <> tail -> Ok(#(char, tail))
+
     _ -> Error(Nil)
   }
 }
 
+// DIGIT (non-zero)    = %x31–39
 fn parse_digit_nz(str: String) -> Result(#(String, String), Nil) {
   case str {
-    "1" as l <> rest
+    "1" as char <> tail
-    | "2" as l <> rest
+    | "2" as char <> tail
-    | "3" as l <> rest
+    | "3" as char <> tail
-    | "4" as l <> rest
+    | "4" as char <> tail
-    | "5" as l <> rest
+    | "5" as char <> tail
-    | "6" as l <> rest
+    | "6" as char <> tail
-    | "7" as l <> rest
+    | "7" as char <> tail
-    | "8" as l <> rest
+    | "8" as char <> tail
-    | "9" as l <> rest -> Ok(#(l, rest))
+    | "9" as char <> tail -> Ok(#(char, tail))
+
     _ -> Error(Nil)
   }
 }
@@ -783,58 +793,59 @@ fn parse_digits(str: String, digits: String) {
 // ALPHA    = %x41–5A | %x61–7A
 fn parse_alpha(str: String) -> Result(#(String, String), Nil) {
   case str {
-    "a" as l <> rest
+    "a" as char <> tail
-    | "b" as l <> rest
+    | "b" as char <> tail
-    | "c" as l <> rest
+    | "c" as char <> tail
-    | "d" as l <> rest
+    | "d" as char <> tail
-    | "e" as l <> rest
+    | "e" as char <> tail
-    | "f" as l <> rest
+    | "f" as char <> tail
-    | "g" as l <> rest
+    | "g" as char <> tail
-    | "h" as l <> rest
+    | "h" as char <> tail
-    | "i" as l <> rest
+    | "i" as char <> tail
-    | "j" as l <> rest
+    | "j" as char <> tail
-    | "k" as l <> rest
+    | "k" as char <> tail
-    | "l" as l <> rest
+    | "l" as char <> tail
-    | "m" as l <> rest
+    | "m" as char <> tail
-    | "n" as l <> rest
+    | "n" as char <> tail
-    | "o" as l <> rest
+    | "o" as char <> tail
-    | "p" as l <> rest
+    | "p" as char <> tail
-    | "q" as l <> rest
+    | "q" as char <> tail
-    | "r" as l <> rest
+    | "r" as char <> tail
-    | "s" as l <> rest
+    | "s" as char <> tail
-    | "t" as l <> rest
+    | "t" as char <> tail
-    | "u" as l <> rest
+    | "u" as char <> tail
-    | "v" as l <> rest
+    | "v" as char <> tail
-    | "w" as l <> rest
+    | "w" as char <> tail
-    | "x" as l <> rest
+    | "x" as char <> tail
-    | "y" as l <> rest
+    | "y" as char <> tail
-    | "z" as l <> rest
+    | "z" as char <> tail
-    | "A" as l <> rest
+    | "A" as char <> tail
-    | "B" as l <> rest
+    | "B" as char <> tail
-    | "C" as l <> rest
+    | "C" as char <> tail
-    | "D" as l <> rest
+    | "D" as char <> tail
-    | "E" as l <> rest
+    | "E" as char <> tail
-    | "F" as l <> rest
+    | "F" as char <> tail
-    | "G" as l <> rest
+    | "G" as char <> tail
-    | "H" as l <> rest
+    | "H" as char <> tail
-    | "I" as l <> rest
+    | "I" as char <> tail
-    | "J" as l <> rest
+    | "J" as char <> tail
-    | "K" as l <> rest
+    | "K" as char <> tail
-    | "L" as l <> rest
+    | "L" as char <> tail
-    | "M" as l <> rest
+    | "M" as char <> tail
-    | "N" as l <> rest
+    | "N" as char <> tail
-    | "O" as l <> rest
+    | "O" as char <> tail
-    | "P" as l <> rest
+    | "P" as char <> tail
-    | "Q" as l <> rest
+    | "Q" as char <> tail
-    | "R" as l <> rest
+    | "R" as char <> tail
-    | "S" as l <> rest
+    | "S" as char <> tail
-    | "T" as l <> rest
+    | "T" as char <> tail
-    | "U" as l <> rest
+    | "U" as char <> tail
-    | "V" as l <> rest
+    | "V" as char <> tail
-    | "W" as l <> rest
+    | "W" as char <> tail
-    | "X" as l <> rest
+    | "X" as char <> tail
-    | "Y" as l <> rest
+    | "Y" as char <> tail
-    | "Z" as l <> rest -> Ok(#(l, rest))
+    | "Z" as char <> tail -> Ok(#(char, tail))
+
     _ -> Error(Nil)
   }
 }

src/gluri/internal/utils.gleam

@@ -109,7 +109,7 @@ pub fn parse_min_max(
   do_parse_min_max(str, "", min, max, parse_fn)
 }
 
-pub fn do_parse_min_max(
+fn do_parse_min_max(
   str: d,
   acc: String,
   min: Int,
@@ -370,30 +370,31 @@ fn unescape_percent(str: String) -> String {
   }
 }
 
-pub fn parse_hex_digit(str) {
+pub fn parse_hex_digit(str: String) -> Result(#(String, String), Nil) {
   case str {
-    "0" as l <> rest
+    "0" as char <> tail
-    | "1" as l <> rest
+    | "1" as char <> tail
-    | "2" as l <> rest
+    | "2" as char <> tail
-    | "3" as l <> rest
+    | "3" as char <> tail
-    | "4" as l <> rest
+    | "4" as char <> tail
-    | "5" as l <> rest
+    | "5" as char <> tail
-    | "6" as l <> rest
+    | "6" as char <> tail
-    | "7" as l <> rest
+    | "7" as char <> tail
-    | "8" as l <> rest
+    | "8" as char <> tail
-    | "9" as l <> rest
+    | "9" as char <> tail
-    | "a" as l <> rest
+    | "a" as char <> tail
-    | "b" as l <> rest
+    | "b" as char <> tail
-    | "c" as l <> rest
+    | "c" as char <> tail
-    | "d" as l <> rest
+    | "d" as char <> tail
-    | "e" as l <> rest
+    | "e" as char <> tail
-    | "f" as l <> rest
+    | "f" as char <> tail
-    | "A" as l <> rest
+    | "A" as char <> tail
-    | "B" as l <> rest
+    | "B" as char <> tail
-    | "C" as l <> rest
+    | "C" as char <> tail
-    | "D" as l <> rest
+    | "D" as char <> tail
-    | "E" as l <> rest
+    | "E" as char <> tail
-    | "F" as l <> rest -> Ok(#(l, rest))
+    | "F" as char <> tail -> Ok(#(char, tail))
+
     _ -> Error(Nil)
   }
 }
@@ -402,14 +403,6 @@ pub fn parse_hex_digits(str, min, max) {
   parse_min_max(str, min, max, parse_hex_digit)
 }
 
-fn encoding_not_needed(i: Int) -> Bool {
-  // $-_.+!*'()
-  case i {
-    36 | 45 | 95 | 46 | 43 | 33 | 42 | 39 | 40 | 41 -> True
-    _ -> False
-  }
-}
-
 fn is_unreserved_char(i: Int) -> Bool {
   case i {
     45 | 46 | 95 | 126 -> True
@@ -659,13 +652,13 @@ pub fn do_percent_encode(str: String) -> String {
 fn encode_codepoint(codepoint: Int) -> String {
   case codepoint <= 127 {
     True -> {
-      case is_unreserved_char(codepoint) || encoding_not_needed(codepoint) {
+      case is_unreserved_char(codepoint) {
         True -> {
           let assert Ok(cpnt) = string.utf_codepoint(codepoint)
           string.from_utf_codepoints([cpnt])
         }
         False -> {
-          "%" <> int.to_base16(codepoint)
+          "%" <> string.pad_start(int.to_base16(codepoint), 2, "0")
         }
       }
     }

test/benchmark.gleam

@@ -1,6 +1,6 @@
+import gleam/string
 import gleam/uri as uri2
 import gluri as uri
-import gluri/internal/parser
 import glychee/benchmark
 import glychee/configuration
 
@@ -10,48 +10,49 @@ pub fn main() {
   configuration.set_pair(configuration.Warmup, 2)
   configuration.set_pair(configuration.Parallel, 2)
 
+  // pop_benchmark()
   parse_benchmark()
   // reg_name_benchmark()
   // ip_benchmark()
 }
 
-@target(erlang)
+// @target(erlang)
-pub fn ip_benchmark() {
+// pub fn ip_benchmark() {
-  benchmark.run(
+//   benchmark.run(
-    [
+//     [
-      benchmark.Function("ip_benchmark", fn(data) {
+//       benchmark.Function("ip_benchmark", fn(data) {
-        fn() {
+//         fn() {
-          let _ = parser.parse_dec_octet(data)
+//           let _ = parser.parse_dec_octet(data)
-          Nil
+//           Nil
-        }
+//         }
-      }),
+//       }),
-    ],
+//     ],
-    [
+//     [
-      benchmark.Data("173", "173"),
+//       benchmark.Data("173", "173"),
-      benchmark.Data("5", "5"),
+//       benchmark.Data("5", "5"),
-      benchmark.Data("200", "200"),
+//       benchmark.Data("200", "200"),
-      benchmark.Data("255", "255"),
+//       benchmark.Data("255", "255"),
-      benchmark.Data("fail", "2b"),
+//       benchmark.Data("fail", "2b"),
-    ],
+//     ],
-  )
+//   )
-}
+// }
 
-@target(erlang)
+// @target(erlang)
-pub fn reg_name_benchmark() {
+// pub fn reg_name_benchmark() {
-  benchmark.run(
+//   benchmark.run(
-    [
+//     [
-      benchmark.Function("reg_name_benchmark", fn(data) {
+//       benchmark.Function("reg_name_benchmark", fn(data) {
-        fn() {
+//         fn() {
-          let _ = parser.parse_reg_name(data)
+//           let _ = parser.parse_reg_name(data)
-          Nil
+//           Nil
-        }
+//         }
-      }),
+//       }),
-    ],
+//     ],
-    [
+//     [
-      benchmark.Data("long", "github.com"),
+//       benchmark.Data("long", "github.com"),
-    ],
+//     ],
-  )
+//   )
-}
+// }
 
 @target(erlang)
 pub fn parse_benchmark() {
@@ -75,7 +76,288 @@ pub fn parse_benchmark() {
         "long",
         "https://github.com/gleam-lang/stdlib/issues/523#issuecomment-3288230480",
       ),
+      benchmark.Data(
+        "with user",
+        "https://test_name:user%20$$$@github.com/gleam-lang/stdlib/issues/523#issuecomment-3288230480",
+      ),
       benchmark.Data("ipv4", "https://192.255.36.4/"),
     ],
   )
 }
+
+@target(erlang)
+pub fn pop_benchmark() {
+  benchmark.run(
+    [
+      benchmark.Function("pop with range", fn(data) { fn() { pop(data, "") } }),
+      benchmark.Function("pop check char", fn(data) { fn() { pop4(data, "") } }),
+      benchmark.Function("pop check result", fn(data) {
+        fn() { pop5(data, "") }
+      }),
+      benchmark.Function("letter as var <> tail", fn(data) {
+        fn() { pop2(data, "") }
+      }),
+      benchmark.Function("letter <> tail", fn(data) { fn() { pop3(data, "") } }),
+    ],
+    [
+      // benchmark.Data("long", "abcdefghijklmnopqrstuvwxyz"),
+      benchmark.Data(
+        "with user",
+        "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ",
+      ),
+      // benchmark.Data("ipv4", "https://192.255.36.4/"),
+    ],
+  )
+}
+
+pub fn pop(input, _) {
+  case string.pop_grapheme(input) {
+    Ok(#(char, tail)) -> {
+      let assert [codepoint] = string.to_utf_codepoints(char)
+      let i = string.utf_codepoint_to_int(codepoint)
+      case i {
+        _ if i >= 0x41 && i <= 0x5A -> pop(tail, char)
+        _ if i >= 0x61 && i <= 0x7A -> pop(tail, char)
+        _ -> Nil
+      }
+    }
+    Error(_) -> Nil
+  }
+}
+
+pub fn pop2(input, _) {
+  case input {
+    "a" as j <> tail
+    | "b" as j <> tail
+    | "c" as j <> tail
+    | "d" as j <> tail
+    | "e" as j <> tail
+    | "f" as j <> tail
+    | "g" as j <> tail
+    | "h" as j <> tail
+    | "i" as j <> tail
+    | "j" as j <> tail
+    | "k" as j <> tail
+    | "l" as j <> tail
+    | "m" as j <> tail
+    | "n" as j <> tail
+    | "o" as j <> tail
+    | "p" as j <> tail
+    | "q" as j <> tail
+    | "r" as j <> tail
+    | "s" as j <> tail
+    | "t" as j <> tail
+    | "u" as j <> tail
+    | "v" as j <> tail
+    | "w" as j <> tail
+    | "x" as j <> tail
+    | "y" as j <> tail
+    | "z" as j <> tail
+    | "A" as j <> tail
+    | "B" as j <> tail
+    | "C" as j <> tail
+    | "D" as j <> tail
+    | "E" as j <> tail
+    | "F" as j <> tail
+    | "G" as j <> tail
+    | "H" as j <> tail
+    | "I" as j <> tail
+    | "J" as j <> tail
+    | "K" as j <> tail
+    | "L" as j <> tail
+    | "M" as j <> tail
+    | "N" as j <> tail
+    | "O" as j <> tail
+    | "P" as j <> tail
+    | "Q" as j <> tail
+    | "R" as j <> tail
+    | "S" as j <> tail
+    | "T" as j <> tail
+    | "U" as j <> tail
+    | "V" as j <> tail
+    | "W" as j <> tail
+    | "X" as j <> tail
+    | "Y" as j <> tail
+    | "Z" as j <> tail -> pop2(tail, j)
+    _ -> Nil
+  }
+}
+
+pub fn pop3(input, _) {
+  case input {
+    "a" <> tail
+    | "b" <> tail
+    | "c" <> tail
+    | "d" <> tail
+    | "e" <> tail
+    | "f" <> tail
+    | "g" <> tail
+    | "h" <> tail
+    | "i" <> tail
+    | "j" <> tail
+    | "k" <> tail
+    | "l" <> tail
+    | "m" <> tail
+    | "n" <> tail
+    | "o" <> tail
+    | "p" <> tail
+    | "q" <> tail
+    | "r" <> tail
+    | "s" <> tail
+    | "t" <> tail
+    | "u" <> tail
+    | "v" <> tail
+    | "w" <> tail
+    | "x" <> tail
+    | "y" <> tail
+    | "z" <> tail
+    | "A" <> tail
+    | "B" <> tail
+    | "C" <> tail
+    | "D" <> tail
+    | "E" <> tail
+    | "F" <> tail
+    | "G" <> tail
+    | "H" <> tail
+    | "I" <> tail
+    | "J" <> tail
+    | "K" <> tail
+    | "L" <> tail
+    | "M" <> tail
+    | "N" <> tail
+    | "O" <> tail
+    | "P" <> tail
+    | "Q" <> tail
+    | "R" <> tail
+    | "S" <> tail
+    | "T" <> tail
+    | "U" <> tail
+    | "V" <> tail
+    | "W" <> tail
+    | "X" <> tail
+    | "Y" <> tail
+    | "Z" <> tail -> pop3(tail, "")
+    _ -> Nil
+  }
+}
+
+pub fn pop4(input, _) {
+  case string.pop_grapheme(input) {
+    Ok(#(char, tail)) -> {
+      case char {
+        "a"
+        | "b"
+        | "c"
+        | "d"
+        | "e"
+        | "f"
+        | "g"
+        | "h"
+        | "i"
+        | "j"
+        | "k"
+        | "l"
+        | "m"
+        | "n"
+        | "o"
+        | "p"
+        | "q"
+        | "r"
+        | "s"
+        | "t"
+        | "u"
+        | "v"
+        | "w"
+        | "x"
+        | "y"
+        | "z"
+        | "A"
+        | "B"
+        | "C"
+        | "D"
+        | "E"
+        | "F"
+        | "G"
+        | "H"
+        | "I"
+        | "J"
+        | "K"
+        | "L"
+        | "M"
+        | "N"
+        | "O"
+        | "P"
+        | "Q"
+        | "R"
+        | "S"
+        | "T"
+        | "U"
+        | "V"
+        | "W"
+        | "X"
+        | "Y"
+        | "Z" -> pop4(tail, char)
+        _ -> Nil
+      }
+    }
+    Error(_) -> Nil
+  }
+}
+
+pub fn pop5(input, _) {
+  case string.pop_grapheme(input) {
+    Ok(#("a" as char, tail))
+    | Ok(#("b" as char, tail))
+    | Ok(#("c" as char, tail))
+    | Ok(#("d" as char, tail))
+    | Ok(#("e" as char, tail))
+    | Ok(#("f" as char, tail))
+    | Ok(#("g" as char, tail))
+    | Ok(#("h" as char, tail))
+    | Ok(#("i" as char, tail))
+    | Ok(#("j" as char, tail))
+    | Ok(#("k" as char, tail))
+    | Ok(#("l" as char, tail))
+    | Ok(#("m" as char, tail))
+    | Ok(#("n" as char, tail))
+    | Ok(#("o" as char, tail))
+    | Ok(#("p" as char, tail))
+    | Ok(#("q" as char, tail))
+    | Ok(#("r" as char, tail))
+    | Ok(#("s" as char, tail))
+    | Ok(#("t" as char, tail))
+    | Ok(#("u" as char, tail))
+    | Ok(#("v" as char, tail))
+    | Ok(#("w" as char, tail))
+    | Ok(#("x" as char, tail))
+    | Ok(#("y" as char, tail))
+    | Ok(#("z" as char, tail))
+    | Ok(#("A" as char, tail))
+    | Ok(#("B" as char, tail))
+    | Ok(#("C" as char, tail))
+    | Ok(#("D" as char, tail))
+    | Ok(#("E" as char, tail))
+    | Ok(#("F" as char, tail))
+    | Ok(#("G" as char, tail))
+    | Ok(#("H" as char, tail))
+    | Ok(#("I" as char, tail))
+    | Ok(#("J" as char, tail))
+    | Ok(#("K" as char, tail))
+    | Ok(#("L" as char, tail))
+    | Ok(#("M" as char, tail))
+    | Ok(#("N" as char, tail))
+    | Ok(#("O" as char, tail))
+    | Ok(#("P" as char, tail))
+    | Ok(#("Q" as char, tail))
+    | Ok(#("R" as char, tail))
+    | Ok(#("S" as char, tail))
+    | Ok(#("T" as char, tail))
+    | Ok(#("U" as char, tail))
+    | Ok(#("V" as char, tail))
+    | Ok(#("W" as char, tail))
+    | Ok(#("X" as char, tail))
+    | Ok(#("Y" as char, tail))
+    | Ok(#("Z" as char, tail)) -> pop4(tail, char)
+    _ -> Nil
+  }
+}

test/benchmark.js

@@ -0,0 +1,16 @@
+import { run, bench, boxplot, summary } from "mitata";
+import { parse } from "../build/dev/javascript/gluri/gluri.mjs";
+import { parse as parse2 } from "../build/dev/javascript/gleam_stdlib/gleam/uri.mjs";
+
+bench("parse", () =>
+  parse(
+    "https://test_name:user%20$$$@github.com/gleam-lang/stdlib/issues/523#issuecomment-3288230480",
+  ),
+);
+bench("parse2", () =>
+  parse2(
+    "https://test_name:user%20$$$@github.com/gleam-lang/stdlib/issues/523#issuecomment-3288230480",
+  ),
+);
+
+await run();

test/gluri_test.gleam

@@ -9,6 +9,66 @@ pub fn main() {
   startest.run(startest.default_config())
 }
 
+pub fn parse_general_tests() {
+  describe("general parsing", [
+    it("mailto parsing", fn() {
+      uri.parse("mailto:Joe@example.com")
+      |> should.equal(Ok(
+        Uri(..empty, scheme: Some("mailto"), path: "Joe@example.com"),
+      ))
+      uri.parse("mailto:Joe@example.com?hello#bye")
+      |> should.equal(Ok(
+        Uri(
+          ..empty,
+          scheme: Some("mailto"),
+          path: "Joe@example.com",
+          query: Some("hello"),
+          fragment: Some("bye"),
+        ),
+      ))
+    }),
+    it("ai gen pass", fn() {
+      let _ = uri.parse("https://example.com") |> should.be_ok
+      let _ =
+        uri.parse("http://www.example.org/resource?id=123&lang=en")
+        |> should.be_ok
+      let _ =
+        uri.parse("ftp://ftp.example.net/pub/files/archive.tar.gz")
+        |> should.be_ok
+      let _ = uri.parse("mailto:user+alias@example.com") |> should.be_ok
+      let _ = uri.parse("urn:isbn:978-3-16-148410-0") |> should.be_ok
+      let _ =
+        uri.parse("ws://socket.example.com:8080/chat?room=42#section2")
+        |> should.be_ok
+      let _ =
+        uri.parse("https://sub.domain.co.uk/path/to/resource/") |> should.be_ok
+      let _ =
+        uri.parse("file:///C:/Windows/System32/drivers/etc/hosts")
+        |> should.be_ok
+      let _ =
+        uri.parse("git+ssh://git@example.com:2222/repo.git") |> should.be_ok
+      let _ =
+        uri.parse(
+          "https://xn--fsqu00a.xn--0zwm56d/%E8%B7%AF%E5%BE%84?%E6%9F%A5%E8%AF%A2=%E5%80%BC#%E7%89%87%E6%AE%B5",
+        )
+        |> should.be_ok
+      Nil
+    }),
+    it("ai gen fail", fn() {
+      let _ = uri.parse("ht!tp://example.com") |> should.be_error
+      let _ = uri.parse("http://exa mple.com") |> should.be_error
+      let _ = uri.parse("://missing-scheme.com") |> should.be_error
+      let _ = uri.parse("http://example.com:80a/") |> should.be_error
+      let _ = uri.parse("http://[2001:db8::1") |> should.be_error
+      let _ = uri.parse("http://example.com/%ZZ") |> should.be_error
+      let _ = uri.parse("http://example.com?%") |> should.be_error
+      let _ = uri.parse("`https://example.com/invalid") |> should.be_error
+      let _ = uri.parse("http://example.com?foo=bar%2") |> should.be_error
+      let _ = uri.parse("http://example.com:12345abc/") |> should.be_error
+    }),
+  ])
+}
+
 pub fn parse_scheme_tests() {
   describe("scheme parsing", [
     it("simple parse", fn() {
@@ -1086,37 +1146,31 @@ pub fn equivalence_tests() {
   ])
 }
 
-const percent_codec_fixtures = [
+const percent_encode_examples = [
-  #(" ", "%20"),
+  #("", ""),
-  #(",", "%2C"),
+  #("%", "%25"),
-  #(";", "%3B"),
+  #("%%", "%25%25"),
-  #(":", "%3A"),
+  #(" \r\n\t\u{B}\f", "%20%0D%0A%09%0B%0C"),
-  #("!", "!"),
+  #(
-  #("?", "%3F"),
+    "-_.~0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ",
-  #("'", "'"),
+    "-_.~0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ",
-  #("(", "("),
+  ),
-  #(")", ")"),
+  #("\u{0}", "%00"),
-  #("[", "%5B"),
+  #("abc\u{00}def", "abc%00def"),
-  #("@", "%40"),
-  #("/", "%2F"),
-  #("\\", "%5C"),
   #("&", "%26"),
-  #("#", "%23"),
+  #(
-  #("=", "%3D"),
+    "!\"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~",
-  #("~", "~"),
+    "%21%22%23%24%25%26%27%28%29%2A%2B%2C-.%2F%3A%3B%3C%3D%3E%3F%40%5B%5C%5D%5E_%60%7B%7C%7D~",
-  #("ñ", "%C3%B1"),
+  ),
-  #("-", "-"),
+  #("“Aha”", "%E2%80%9CAha%E2%80%9D"),
-  #("_", "_"),
+  #("\u{201C}Aha\u{201D}", "%E2%80%9CAha%E2%80%9D"),
-  #(".", "."),
+  #("*+,=>/", "%2A%2B%2C%3D%3E%2F"),
-  #("*", "*"),
-  #("+", "+"),
-  #("100% great+fun", "100%25%20great+fun"),
 ]
 
 pub fn percent_encode_tests() {
   describe("percent encoding", [
     it("encoding", fn() {
-      percent_codec_fixtures
+      percent_encode_examples
       |> list.map(fn(t) {
         let #(a, b) = t
         uri.percent_encode(a)
@@ -1125,7 +1179,7 @@ pub fn percent_encode_tests() {
       Nil
     }),
     it("decoding", fn() {
-      percent_codec_fixtures
+      percent_encode_examples
       |> list.map(fn(t) {
         let #(a, b) = t
         uri.percent_decode(b)