xberg 1.3.4 → 1.3.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 776ad3d215ba3a5adf1325330a593b9b4853acc77504fdc8d11161dea9374d6d
-  data.tar.gz: d17b3122b02e8420d6b7d46e374278b6f1d9fde0889c9543f63cae2ef9a17f29
+  metadata.gz: 2fb31147a3b3ff05c73459891b1c38bd0b056ba12054ee443ebbb1dee6dea9c9
+  data.tar.gz: 8e0c0a12963f1a9cf6e894b72638040f2e751ff85fcd6642aff75bf61d1b1246
 SHA512:
-  metadata.gz: fb7cdedf5123544de3abaa28010bb47127ce0fdb389b95d1ebeb6d2920c6c2dff92bd2d92081f2dad46659e6164cae092fc85dac955dca0c94de87bd3724cef9
-  data.tar.gz: e45ffd440c69abf6ed32e28d945e5a08bc772d4253d542284248210dc5c2ea70efd018386b4614d827fc889756381d50e72b2b32dc9f82e00b44d5f965534721
+  metadata.gz: b99cac72c23ba63bccbc846f30ffa1eac13c4b75f6b5812b6cc050c8e3ba6ea82c93c153ce4b74e883220d869e55bce44541e299cdc90c86af6c4b63d1ba87be
+  data.tar.gz: 18e9fc96db580bc7bafd9578eaca72ad2f2d546245e186781426c9c3707294a237b576100b88cd38be60928a1166fe79f63574003b2f368dca2db7abfe0c5f41

data/ext/xberg_rb/native/Cargo.lock

@@ -167,9 +167,6 @@ name = "arbitrary"
 version = "1.4.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "c3d036a3c4ab069c7b410a2ce876bd74808d2d0888a82667669f8e783a898bf1"
-dependencies = [
- "derive_arbitrary",
-]
 
 [[package]]
 name = "arc-swap"
@@ -707,18 +704,18 @@ dependencies = [
 
 [[package]]
 name = "bon"
-version = "3.10.1"
+version = "3.10.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "60eafe0d77c3a2fc292c1d1346c3041b33c0a108085a2afabf672b70f69dbbc9"
+checksum = "214f5df094ce551a10a30ffb9c70243d61f121a3d985a6495933e181dee6a7d2"
 dependencies = [
  "bon-macros",
 ]
 
 [[package]]
 name = "bon-macros"
-version = "3.10.1"
+version = "3.10.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "bd0f9631d8aaaee112c41985d675ef269e02acbd4f33122836af4f0c5f699ff6"
+checksum = "2706da6c749998cc555d04184909608956a69ad3f8ab1a076fbc867b4a3c3ce5"
 dependencies = [
  "darling 0.24.1",
  "ident_case",
@@ -899,7 +896,7 @@ dependencies = [
  "log",
  "quick-xml 0.41.0",
  "serde",
- "zip 8.6.0",
+ "zip",
 ]
 
 [[package]]
@@ -925,7 +922,7 @@ dependencies = [
  "tokenizers 0.22.2",
  "yoke",
  "zerocopy",
- "zip 8.6.0",
+ "zip",
 ]
 
 [[package]]
@@ -1001,9 +998,9 @@ dependencies = [
 
 [[package]]
 name = "cc"
-version = "1.5.1"
+version = "1.6.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f360145194ee8e21db5ee7f3fcd4fe52210864c75c985dae33218202c8bbe040"
+checksum = "f74872d07caf508b30a21f6836e7d7016a2eaf7d9ff4f48deaa58cd8a0407630"
 dependencies = [
  "find-msvc-tools",
  "jobserver",
@@ -1408,9 +1405,9 @@ dependencies = [
 
 [[package]]
 name = "crawlberg"
-version = "1.8.0"
+version = "1.10.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5c2db0cf209917529c36a2238d1f57b01c30a170bc3d3c29828dbbd8daca39fb"
+checksum = "5deba1d98a2443b6c43beca9ec699ed3bb2cd1a19d010bdebaceb4a57f119ce4"
 dependencies = [
  "ahash",
  "aho-corasick",
@@ -1420,8 +1417,10 @@ dependencies = [
  "base64 0.23.1",
  "blake3",
  "chrono",
+ "crawlberg-robots",
  "dashmap",
  "encoding_rs",
+ "fancy-regex 0.19.2",
  "flate2",
  "futures",
  "getrandom 0.3.4",
@@ -1429,15 +1428,22 @@ dependencies = [
  "html-escape",
  "html-to-markdown-rs",
  "html5ever",
+ "hyper-util",
+ "icu_casemap",
+ "icu_normalizer",
+ "idna",
  "ipnet",
  "liter-llm",
  "memchr",
  "notify",
  "opentelemetry",
  "opentelemetry-semantic-conventions",
+ "percent-encoding",
  "quick-xml 0.42.0",
  "regex",
+ "regex-syntax",
  "reqwest",
+ "same-file",
  "serde",
  "serde_json",
  "sha2 0.11.0",
@@ -1451,6 +1457,12 @@ dependencies = [
  "web-time",
 ]
 
+[[package]]
+name = "crawlberg-robots"
+version = "1.10.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d8499344c8bacd91c1cd16c36e4998769ecc1210179e88d2b80eb33468d6fee9"
+
 [[package]]
 name = "crc32fast"
 version = "1.5.2"
@@ -1688,17 +1700,6 @@ version = "0.5.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "7cd812cc2bc1d69d4764bd80df88b4317eaef9e773c75226407d9bc0876b211c"
 
-[[package]]
-name = "derive_arbitrary"
-version = "1.4.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1e567bd82dcff979e4b03460c307b3cdc9e96fde3d73bed1496d2bc75d9dd62a"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn 2.0.119",
-]
-
 [[package]]
 name = "derive_builder"
 version = "0.20.2"
@@ -2105,9 +2106,9 @@ dependencies = [
 
 [[package]]
 name = "fearless_simd"
-version = "0.4.1"
+version = "1.0.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b97b65636e5b9ef369943878ac74335ba1c55c1cb6adbf1e2c293c624248d693"
+checksum = "f3772b63c40606beea8fe1f7b06a60de27ceb9bfc2d9dad3427368ca41dc7e62"
 
 [[package]]
 name = "filetime"
@@ -2191,9 +2192,9 @@ checksum = "77ce24cb58228fbb8aa041425bb1050850ac19177686ea6e0f41a70416f56fdb"
 
 [[package]]
 name = "font-types"
-version = "0.12.5"
+version = "0.12.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b8eb065f3251655b3c90e22e5e363f310fc5332fb3402e37bbc94752283248f6"
+checksum = "6c92d9081c7ab3bd83ef1eecdfad090f3d966e3ce74106779664114cfe1a6274"
 dependencies = [
  "bytemuck",
 ]
@@ -2834,24 +2835,24 @@ dependencies = [
 
 [[package]]
 name = "hayro-ccitt"
-version = "0.3.0"
+version = "0.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9f4d0e94ddd48749f06bbe4e5389fb9799a0c45bcaf00495042076ef05e3241a"
+checksum = "2866f0d976e76497480b853d5e258dab6e4b89ecfe991124e08f15dfa8fd03ad"
 
 [[package]]
 name = "hayro-jbig2"
-version = "0.3.0"
+version = "0.3.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "69374b3668dd45aeb3d3145cda68f2c7b4f223aaa2511e67d076f1c7d741388d"
+checksum = "000e025850f4592ddcaee85640db70704151f2bafd43b6f431aa5b6a2c6a35a9"
 dependencies = [
  "hayro-ccitt",
 ]
 
 [[package]]
 name = "hayro-jpeg2000"
-version = "0.4.0"
+version = "0.4.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "bab0c77d09c0d9b144d429d0bf0fcd4c63c01d5f6c33f9b8ed501283e0f1ef76"
+checksum = "122c5f834c37a2260b9c851bb8be685cf539b2a8b94b704a931ef7d2525d5835"
 dependencies = [
  "fearless_simd",
 ]
@@ -2957,9 +2958,9 @@ checksum = "c9356095b4b41197bba32173600e1582792cda618f65d12f68e2e77d273413c5"
 
 [[package]]
 name = "html-to-markdown-rs"
-version = "3.16.0"
+version = "3.17.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "74c4d770081ede1c6ec100e61a57cac65bf41d28ab5516b1eae907b9f3415778"
+checksum = "1018d58b647013e9968c46a6625321611c9026df6a44e12cea4c83b207c5e2bb"
 dependencies = [
  "ahash",
  "astral-tl",
@@ -3154,6 +3155,28 @@ dependencies = [
  "cc",
 ]
 
+[[package]]
+name = "icu_casemap"
+version = "2.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f2edfd120d8445de853cf53a1abec916f638f933a70a6f4bf8f7e2d7b7f884e9"
+dependencies = [
+ "icu_casemap_data",
+ "icu_collections",
+ "icu_locale_core",
+ "icu_properties",
+ "icu_provider",
+ "potential_utf",
+ "writeable",
+ "zerovec",
+]
+
+[[package]]
+name = "icu_casemap_data"
+version = "2.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "059fcda0fc8d33effc6633a8cf0d8a3c17caf7d7c91918fba9918744f501e4e8"
+
 [[package]]
 name = "icu_collections"
 version = "2.3.0"
@@ -3162,6 +3185,7 @@ checksum = "fa68d21081c4a05d5a901a1c62add574c77048b6a1c67be3b50ce0b60d4ca513"
 dependencies = [
  "displaydoc",
  "potential_utf",
+ "serde",
  "utf8_iter",
  "yoke",
  "zerofrom",
@@ -3621,9 +3645,9 @@ dependencies = [
 
 [[package]]
 name = "jsonschema"
-version = "0.58.4"
+version = "0.58.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5761b0ca0a31464030d7c2a98087fde3124ebb5e09b5a112cdb815f03b8ee150"
+checksum = "ea18b8d5e1469b1bdd6b349169305141316a7c26da27f239773b65c830c0f5d6"
 dependencies = [
  "ahash",
  "bytecount",
@@ -3649,18 +3673,18 @@ dependencies = [
 
 [[package]]
 name = "jsonschema-regex"
-version = "0.58.4"
+version = "0.58.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f1ffba4378a3d6ee7f1ab65a8ed95c711ccfe4a3870932bdce89d8260abed38a"
+checksum = "854e9e22c420535035e95eed240cbfb4a5f4b7da74d122d63144e0c31926dbfd"
 dependencies = [
  "regex-syntax",
 ]
 
 [[package]]
 name = "jsonschema-value"
-version = "0.58.4"
+version = "0.58.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "386121f898b560ff6aaa7974952720e9d9276a201eca31296387c2ed8707855e"
+checksum = "4e3bfac11ac357ec620880025757d38c67da8fd5b9ef5024d19f0f360002b4ca"
 dependencies = [
  "ahash",
  "bytecount",
@@ -4159,15 +4183,14 @@ dependencies = [
 
 [[package]]
 name = "mel_spec"
-version = "0.4.0"
+version = "0.5.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0cfc86c477482e75cca77856f89cd5cb485c91331269abff41b9b15364a49492"
+checksum = "53674cd00346021f4d45f82c8645986e9aef1dc188f3a98a5cf7ebaa75a5b659"
 dependencies = [
  "cc",
  "image 0.24.9",
  "ndarray 0.16.1",
- "ndarray-npy",
- "num",
+ "realfft",
  "rustfft",
 ]
 
@@ -4253,9 +4276,9 @@ dependencies = [
 
 [[package]]
 name = "mio"
-version = "1.2.3"
+version = "1.2.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4b18443e9c262bfe8fa82f51666e2642c53393f7e5c27b3e1aeab922cff5b9d8"
+checksum = "1788edb87fdc09c7e26304471e2f5be8cdefb1b6930d6e3985fc02ff53bf86ee"
 dependencies = [
  "libc",
  "log",
@@ -4435,20 +4458,6 @@ dependencies = [
  "rawpointer",
 ]
 
-[[package]]
-name = "ndarray-npy"
-version = "0.9.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7b313788c468c49141a9d9b6131fc15f403e6ef4e8446a0b2e18f664ddb278a9"
-dependencies = [
- "byteorder",
- "ndarray 0.16.1",
- "num-complex",
- "num-traits",
- "py_literal",
- "zip 2.4.2",
-]
-
 [[package]]
 name = "new_debug_unreachable"
 version = "1.0.6"
@@ -4953,48 +4962,6 @@ version = "2.3.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "9b4f627cb1b25917193a259e49bdad08f671f8d9708acfd5fe0a8c1455d87220"
 
-[[package]]
-name = "pest"
-version = "2.9.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "45d3aca230fad2e6f6317ca0a72724338c4960cb97168a85cdee66df4a9a21a8"
-dependencies = [
- "memchr",
- "ucd-trie",
-]
-
-[[package]]
-name = "pest_derive"
-version = "2.9.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "284b60557f2c4a2e72ad3f2d34d42685a2fa4a6a61d0d2a10c0ae2a5e916c2cf"
-dependencies = [
- "pest",
- "pest_generator",
-]
-
-[[package]]
-name = "pest_generator"
-version = "2.9.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1d9d1f08a115309ee99268cf85e5228e0e56aa9caf8841ec12866b6be07c3109"
-dependencies = [
- "pest",
- "pest_meta",
- "proc-macro2",
- "quote",
- "syn 2.0.119",
-]
-
-[[package]]
-name = "pest_meta"
-version = "2.9.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ed93ba1a9ffcca32130a5188701c81c0c49cf00d4b7c5007d5148951d743adcb"
-dependencies = [
- "pest",
-]
-
 [[package]]
 name = "phc"
 version = "0.6.1"
@@ -5221,9 +5188,9 @@ dependencies = [
 
 [[package]]
 name = "powerfmt"
-version = "0.2.0"
+version = "0.2.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "439ee305def115ba05938db6eb1644ff94165c5ab5e9420d1c1bcedbba909391"
+checksum = "4a6394b9e965e73d0a289ee54f589087e2c676aedf60885baf52c76b771e4958"
 
 [[package]]
 name = "ppmd-rust"
@@ -5375,19 +5342,6 @@ version = "0.1.30"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d55d956fa96f5ec02be2e13af0e20391a5aa83d6a074e3ad368959d0fab299ea"
 
-[[package]]
-name = "py_literal"
-version = "0.4.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "102df7a3d46db9d3891f178dcc826dc270a6746277a9ae6436f8d29fd490a8e1"
-dependencies = [
- "num-bigint",
- "num-complex",
- "num-traits",
- "pest",
- "pest_derive",
-]
-
 [[package]]
 name = "qcms"
 version = "0.3.0"
@@ -5772,6 +5726,15 @@ dependencies = [
  "once_cell",
 ]
 
+[[package]]
+name = "realfft"
+version = "3.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f821338fddb99d089116342c46e9f1fbf3828dba077674613e734e01d6ea8677"
+dependencies = [
+ "rustfft",
+]
+
 [[package]]
 name = "reborrow"
 version = "0.5.5"
@@ -5828,9 +5791,9 @@ dependencies = [
 
 [[package]]
 name = "referencing"
-version = "0.58.4"
+version = "0.58.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5cfa1c91745dbca76a406b926918d7c63e15a4525ebab939fec3e015a1fb82b1"
+checksum = "590efadb0a669f1712c1e3ab810a55d4c3d20d0127c455f7236dd0710aba0346"
 dependencies = [
  "ahash",
  "fluent-uri",
@@ -6522,9 +6485,9 @@ dependencies = [
 
 [[package]]
 name = "serde_toon_format"
-version = "0.1.2"
+version = "0.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "45abc88be1b465c1676fbbe94f866e872d3e9c22b3c10de6e1d90ae76dd45d51"
+checksum = "83047917cd35d3cf33b1cfa3085e048546c8fb0f1f86ab90aa8e680196679fc3"
 dependencies = [
  "itoa",
  "memchr",
@@ -7538,9 +7501,9 @@ dependencies = [
 
 [[package]]
 name = "tokio"
-version = "1.53.1"
+version = "1.53.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "202caea871b69668250d242070849eb495be178ed697a3e98aebce5bc81a0bed"
+checksum = "e95f91fcc7a621e8b030f6aa23c71fe9838ae2fb4d8118b75602a328f5144044"
 dependencies = [
  "bytes",
  "libc",
@@ -7961,9 +7924,9 @@ checksum = "ca0d1bf6fdd806e43ae5198f82f527056d359def39e54e67a0f478ac09dac081"
 
 [[package]]
 name = "tree-sitter-language-pack"
-version = "1.20.0"
+version = "1.21.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a360d1c9fd92fc96d9802ce77422d1a90d59d2ec767ef2c0833691f9ac86689c"
+checksum = "1ad8597fc66b66d5ba13729827f07f47b89f9f8336b7441c030cad5ecb63969f"
 dependencies = [
  "ahash",
  "cc",
@@ -7992,9 +7955,9 @@ checksum = "e421abadd41a4225275504ea4d6566923418b7f05506fbc9c0fe86ba7396114b"
 
 [[package]]
 name = "twox-hash"
-version = "2.1.4"
+version = "2.1.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5283634e518fe9e82c7b20520bb4bc209009fd16c82077c802f8111ecbb0117a"
+checksum = "86a801b3cea342a06d468c8710662aa29e5e05e4f5c0d62f00bbb7f2ad7941c2"
 
 [[package]]
 name = "typed-arena"
@@ -8068,17 +8031,11 @@ dependencies = [
  "unicode-math-class",
 ]
 
-[[package]]
-name = "ucd-trie"
-version = "0.1.7"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2896d95c02a80c6d6a5d6e953d479f5ddf2dfdb6a244441010e373ac0fb88971"
-
 [[package]]
 name = "unhwp"
-version = "0.13.0"
+version = "0.13.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c58f5bccc79e6b61afb38e13497a5eefcf6d47d9279f6fe2955f2352386c5780"
+checksum = "b7c1ba4cd5dbec145f52c85932321667c164ed71852a13f5ebc5e46182b7e4a1"
 dependencies = [
  "cfb 0.15.0",
  "flate2",
@@ -8092,14 +8049,14 @@ dependencies = [
  "thiserror 2.0.21",
  "unicode-normalization",
  "unparser-shared",
- "zip 8.6.0",
+ "zip",
 ]
 
 [[package]]
 name = "unicase"
-version = "2.9.0"
+version = "2.10.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "dbc4bc3a9f746d862c45cb89d705aa10f187bb96c76001afab07a0d35ce60142"
+checksum = "357cc3acc6a036009fd6c973ed009037c732d60d0b4f6c673e9041497482a28f"
 
 [[package]]
 name = "unicode-bidi"
@@ -8331,9 +8288,9 @@ dependencies = [
 
 [[package]]
 name = "uuid"
-version = "1.26.1"
+version = "1.27.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2ef6dac1e96601b4fb3acccccff2139741fcb757cb9a36089bf5be91cfb285ce"
+checksum = "97277d36b9c3ace13e58fa6e753f8b0bbbf302a18dd193240d70f9e29681059a"
 dependencies = [
  "getrandom 0.4.3",
  "js-sys",
@@ -8999,9 +8956,9 @@ dependencies = [
 
 [[package]]
 name = "xberg"
-version = "1.3.4"
+version = "1.3.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ca1840b552a16d2d9a2dbb9dec557e822cc9c834b32c6740726413d508e463e2"
+checksum = "31c148cfc7f6f51d18e1c8368ff14d9cebb4881464ffb8c7aa856e581179b41f"
 dependencies = [
  "aes-gcm",
  "ahash",
@@ -9133,14 +9090,14 @@ dependencies = [
  "xberg-pdfium-render",
  "xberg-tesseract",
  "zeroize",
- "zip 8.6.0",
+ "zip",
 ]
 
 [[package]]
 name = "xberg-candle-ocr"
-version = "1.3.4"
+version = "1.3.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c1fcf7160e4c88f64b30fef0cca75c849c41874fce7269066a152a45f008fe80"
+checksum = "cda49ad9614d9ac56ce0e83ea13315af2219ffce77e9e581c8029f38351a3614"
 dependencies = [
  "base64 0.23.1",
  "candle-core",
@@ -9160,9 +9117,9 @@ dependencies = [
 
 [[package]]
 name = "xberg-gliner"
-version = "1.3.4"
+version = "1.3.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d1ecc238993ea3b1f38084c3fbe41a689ed8321145454078057c5991522d02e9"
+checksum = "cbb95e848402e644ba5044c1ed2dbc05f482c8e2f9d531fc44417c9e0d44fbd2"
 dependencies = [
  "getrandom 0.2.17",
  "getrandom 0.3.4",
@@ -9176,9 +9133,9 @@ dependencies = [
 
 [[package]]
 name = "xberg-libheif"
-version = "1.3.4"
+version = "1.3.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "db2f00105c1a2eaa09437849281da3e917410cf3dd45fbc8782bbc96d2eacde6"
+checksum = "22a1a4e50e9e39d2dbbba80d1c5b789e366eae939abe54e925fbbecac487a5ec"
 dependencies = [
  "cfg-if",
  "enumn",
@@ -9189,9 +9146,9 @@ dependencies = [
 
 [[package]]
 name = "xberg-libwpd"
-version = "1.3.4"
+version = "1.3.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2ee648b64ea792f80741925552acd13e69bf771d9b86039555e660593d6732c0"
+checksum = "f70f597ee8b0dbe48703f35b6c1c9723b47ebe53893ff132836d24cfdd9fc141"
 dependencies = [
  "cc",
  "flate2",
@@ -9204,9 +9161,9 @@ dependencies = [
 
 [[package]]
 name = "xberg-native-pdf"
-version = "1.3.4"
+version = "1.3.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4fe79b781a7212508ab39b2be14dcde450e006e2b288207ac6f3eb130804e6bc"
+checksum = "baaee7fbbed92c1b495f5b6d8fc0248710916457ce76b539017d158b3520abaa"
 dependencies = [
  "aes 0.9.3",
  "base64 0.23.1",
@@ -9251,9 +9208,9 @@ dependencies = [
 
 [[package]]
 name = "xberg-paddle-ocr"
-version = "1.3.4"
+version = "1.3.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f5ff67ca24c1f2a22730c9e6dbcd5f5614b244fe8fb74c3e1e9b13073f8a4c9d"
+checksum = "5610a02a608e2a154948b5ff87126f703e82045e6512fa23d5dcbdeac8c2ff0c"
 dependencies = [
  "geo-clipper",
  "geo-types",
@@ -9268,9 +9225,9 @@ dependencies = [
 
 [[package]]
 name = "xberg-pdfium-render"
-version = "1.3.4"
+version = "1.3.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "90013a4daf6eca014fd0e1fb9cea7cbac5ba93316338a4181feead9aea2b1f78"
+checksum = "4a9cf0638be244877d99b2ae6134ba1a88b7e16ec9e15c82bb3a5c0a1f063cb8"
 dependencies = [
  "bitflags 2.13.2",
  "bytemuck",
@@ -9287,7 +9244,7 @@ dependencies = [
 
 [[package]]
 name = "xberg-rb"
-version = "1.3.4"
+version = "1.3.5"
 dependencies = [
  "async-trait",
  "crawlberg",
@@ -9303,9 +9260,9 @@ dependencies = [
 
 [[package]]
 name = "xberg-tesseract"
-version = "1.3.4"
+version = "1.3.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "29967431e67fd43c971d9d348935ee1ca58223bb3110b470dca7fdda0fd67c30"
+checksum = "611d0b8d32a011468a56fa123475d6a12e5f290056c5259fa694dc82622cab89"
 dependencies = [
  "cc",
  "cmake",
@@ -9315,7 +9272,7 @@ dependencies = [
  "sha2 0.11.0",
  "tempfile",
  "thiserror 2.0.21",
- "zip 8.6.0",
+ "zip",
 ]
 
 [[package]]
@@ -9590,23 +9547,6 @@ dependencies = [
  "syn 3.0.6",
 ]
 
-[[package]]
-name = "zip"
-version = "2.4.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "fabe6324e908f85a1c52063ce7aa26b68dcb7eb6dbc83a2d148403c9bc3eba50"
-dependencies = [
- "arbitrary",
- "crc32fast",
- "crossbeam-utils",
- "displaydoc",
- "flate2",
- "indexmap",
- "memchr",
- "thiserror 2.0.21",
- "zopfli",
-]
-
 [[package]]
 name = "zip"
 version = "8.6.0"

data/ext/xberg_rb/native/Cargo.toml

@@ -1,11 +1,11 @@
 # This file is auto-generated by alef — DO NOT EDIT.
-# alef:hash:715ad8642ffff546e63c46773a283e4a03bcb5b87f1ccadddc688baa7e1c4d1d
+# alef:hash:d62e9addf1ad2582f441641d4a00616cca3d0d9f70027e36f6683c9a097c42d2
 # To regenerate: alef generate
 # To verify freshness: alef verify
 
 [package]
 name = "xberg-rb"
-version = "1.3.4"
+version = "1.3.5"
 edition = "2024"
 license = "MIT"
 description = "High-performance document intelligence library"
@@ -72,8 +72,8 @@ xml = ["xberg/xml"]
 
 [dependencies]
 async-trait = "0.1"
-crawlberg = { default-features = false, version = "1.8.0" }
-html-to-markdown-rs = { default-features = false, version = "3.16" }
+crawlberg = { default-features = false, version = "1.10.1" }
+html-to-markdown-rs = { default-features = false, version = "3.17.1" }
 magnus = "0.9"
 rb-sys = ">=0.9.130, <0.10"
 serde = { version = "1", features = ["derive"] }
@@ -82,13 +82,13 @@ tokio = { version = "1", features = ["rt-multi-thread"] }
 tracing = "0.1"
 
 [target.'cfg(all(target_os = "macos", target_arch = "x86_64"))'.dependencies]
-xberg = { version = "1.3.4", default-features = false, features = ["macos-intel-target", "simd-utf8"] }
+xberg = { version = "1.3.5", default-features = false, features = ["macos-intel-target", "simd-utf8"] }
 
 [target.'cfg(not(any(target_os = "windows", all(target_os = "macos", target_arch = "x86_64"))))'.dependencies]
-xberg = { version = "1.3.4", features = ["full", "pdf", "pdf-pdfium", "ocr", "paddle-ocr", "paddle-ocr-types", "layout-detection", "layout-types", "embeddings", "embedding-presets", "reranker", "reranker-presets", "sparse-embeddings", "sparse-embedding-presets", "late-interaction", "late-interaction-presets", "chunking", "keywords-yake", "keywords-rake", "language-detection", "html", "tree-sitter", "office", "email", "archives", "stopwords", "auto-rotate", "auto-rotate-types", "tokio-runtime", "api", "api-types", "mcp", "liter-llm", "quality", "svg", "transcription", "transcription-types", "classification", "captioning", "formula-recognition", "ner-onnx", "ner-llm", "diff", "markdown-footnotes", "redaction", "enrichment", "heuristics", "presets", "structured", "url-ingestion"] }
+xberg = { version = "1.3.5", features = ["full", "pdf", "pdf-pdfium", "ocr", "paddle-ocr", "paddle-ocr-types", "layout-detection", "layout-types", "embeddings", "embedding-presets", "reranker", "reranker-presets", "sparse-embeddings", "sparse-embedding-presets", "late-interaction", "late-interaction-presets", "chunking", "keywords-yake", "keywords-rake", "language-detection", "html", "tree-sitter", "office", "email", "archives", "stopwords", "auto-rotate", "auto-rotate-types", "tokio-runtime", "api", "api-types", "mcp", "liter-llm", "quality", "svg", "transcription", "transcription-types", "classification", "captioning", "formula-recognition", "ner-onnx", "ner-llm", "diff", "markdown-footnotes", "redaction", "enrichment", "heuristics", "presets", "structured", "url-ingestion"] }
 
 [target.'cfg(target_os = "windows")'.dependencies]
-xberg = { version = "1.3.4", default-features = false, features = ["windows-gnu-target", "simd-utf8"] }
+xberg = { version = "1.3.5", default-features = false, features = ["windows-gnu-target", "simd-utf8"] }
 
 # This crate deliberately does not use `[lints]` / `workspace = true`: its C-ABI /
 # PyO3 / napi / ext-php-rs / NIF boundary requires `unsafe` code, and the workspace's

data/ext/xberg_rb/src/lib.rs

@@ -1,5 +1,5 @@
 // This file is auto-generated by alef. DO NOT EDIT.
-// alef:hash:3e862eca7b1250c4aa8d40b5563353572b94bd205d31f9f57c9efc950fcd26cb
+// alef:hash:3751950d58564a09b72eb0e65f2f992e39f3dcd2cd851cd4493626ba6caef9a0
 // Re-generate with: alef generate
 #![allow(dead_code, unused_imports, unused_variables)]
 #![allow(
@@ -1126,7 +1126,7 @@ impl BoundingBox {
     }
 }
 
-#[derive(Clone, Debug, serde::Serialize)]
+#[derive(Clone, serde::Serialize)]
 #[serde(default)]
 #[magnus::wrap(class = "Xberg::BrowserConfig")]
 pub struct BrowserConfig {
@@ -1146,6 +1146,8 @@ pub struct BrowserConfig {
     robots_user_agent: Option<String>,
     capture_network_events: bool,
     session_affinity: bool,
+    chrome_path: Option<String>,
+    chrome_args: Vec<String>,
 }
 
 unsafe impl IntoValueFromNative for BrowserConfig {}
@@ -1175,6 +1177,29 @@ impl magnus::TryConvert for BrowserConfig {
 }
 
 unsafe impl TryConvertOwned for BrowserConfig {}
+impl std::fmt::Debug for BrowserConfig {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        let mut debug = f.debug_struct("BrowserConfig");
+        debug.field("mode", &self.mode);
+        debug.field("backend", &self.backend);
+        debug.field("endpoint", &"<redacted>");
+        debug.field("timeout", &self.timeout);
+        debug.field("overall_timeout", &self.overall_timeout);
+        debug.field("shutdown_timeout", &self.shutdown_timeout);
+        debug.field("wait", &self.wait);
+        debug.field("wait_selector", &self.wait_selector);
+        debug.field("extra_wait", &self.extra_wait);
+        debug.field("proxy", &self.proxy);
+        debug.field("block_url_patterns", &self.block_url_patterns);
+        debug.field("eval_script", &"<redacted>");
+        debug.field("robots_user_agent", &self.robots_user_agent);
+        debug.field("capture_network_events", &self.capture_network_events);
+        debug.field("session_affinity", &self.session_affinity);
+        debug.field("chrome_path", &self.chrome_path);
+        debug.field("chrome_args", &"<redacted>");
+        debug.finish()
+    }
+}
 impl<'de> serde::Deserialize<'de> for BrowserConfig {
     fn deserialize<D>(deserializer: D) -> ::core::result::Result<Self, D::Error>
     where
@@ -1332,6 +1357,24 @@ impl BrowserConfig {
                 })?,
                 None => true,
             },
+            chrome_path: match kwargs.get(ruby.to_symbol("chrome_path")).filter(|v| !v.is_nil()) {
+                Some(v) => Some(String::try_convert(v).map_err(|e| {
+                    magnus::Error::new(
+                        unsafe { magnus::Ruby::get_unchecked() }.exception_type_error(),
+                        format!("invalid value for `chrome_path`: {}", e),
+                    )
+                })?),
+                None => None,
+            },
+            chrome_args: match kwargs.get(ruby.to_symbol("chrome_args")) {
+                Some(v) => <Vec<String>>::try_convert(v).map_err(|e| {
+                    magnus::Error::new(
+                        unsafe { magnus::Ruby::get_unchecked() }.exception_type_error(),
+                        format!("invalid value for `chrome_args`: {}", e),
+                    )
+                })?,
+                None => Default::default(),
+            },
         })
     }
 
@@ -1394,6 +1437,14 @@ impl BrowserConfig {
     fn session_affinity(&self) -> bool {
         self.session_affinity
     }
+
+    fn chrome_path(&self) -> Option<String> {
+        self.chrome_path.clone()
+    }
+
+    fn chrome_args(&self) -> Vec<String> {
+        self.chrome_args.clone()
+    }
 }
 
 #[derive(Clone, Debug, serde::Serialize, serde::Deserialize)]
@@ -4090,6 +4141,7 @@ pub struct ConversionOptions {
     include_document_structure: bool,
     extract_images: bool,
     max_image_size: u64,
+    max_input_size: Option<u64>,
     capture_svg: bool,
     infer_dimensions: bool,
     max_depth: Option<usize>,
@@ -4490,6 +4542,15 @@ impl ConversionOptions {
                 })?,
                 None => 5242880,
             },
+            max_input_size: match kwargs.get(ruby.to_symbol("max_input_size")).filter(|v| !v.is_nil()) {
+                Some(v) => Some(u64::try_convert(v).map_err(|e| {
+                    magnus::Error::new(
+                        unsafe { magnus::Ruby::get_unchecked() }.exception_type_error(),
+                        format!("invalid value for `max_input_size`: {}", e),
+                    )
+                })?),
+                None => None,
+            },
             capture_svg: match kwargs.get(ruby.to_symbol("capture_svg")) {
                 Some(v) => bool::try_convert(v).map_err(|e| {
                     magnus::Error::new(
@@ -4703,6 +4764,10 @@ impl ConversionOptions {
         self.max_image_size
     }
 
+    fn max_input_size(&self) -> Option<u64> {
+        self.max_input_size
+    }
+
     fn capture_svg(&self) -> bool {
         self.capture_svg
     }
@@ -4995,7 +5060,7 @@ impl CoreProperties {
     }
 }
 
-#[derive(Clone, Debug, serde::Serialize, serde::Deserialize)]
+#[derive(Clone, serde::Serialize, serde::Deserialize)]
 #[serde(default)]
 #[magnus::wrap(class = "Xberg::CrawlConfig")]
 pub struct CrawlConfig {
@@ -5015,6 +5080,7 @@ pub struct CrawlConfig {
     include_paths: Vec<String>,
     exclude_paths: Vec<String>,
     path_patterns_match_query: bool,
+    path_patterns_match_url: bool,
     dedup_include_query: bool,
     strip_tracking_params: bool,
     tracking_params: Vec<String>,
@@ -5083,6 +5149,67 @@ impl magnus::TryConvert for CrawlConfig {
 }
 
 unsafe impl TryConvertOwned for CrawlConfig {}
+impl std::fmt::Debug for CrawlConfig {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        let mut debug = f.debug_struct("CrawlConfig");
+        debug.field("max_depth", &self.max_depth);
+        debug.field("max_pages", &self.max_pages);
+        debug.field("max_links_per_page", &self.max_links_per_page);
+        debug.field("max_concurrent", &self.max_concurrent);
+        debug.field("crawl_strategy", &self.crawl_strategy);
+        debug.field("content_filter", &self.content_filter);
+        debug.field("bm25_query", &self.bm25_query);
+        debug.field("bm25_threshold", &self.bm25_threshold);
+        debug.field("respect_robots_txt", &self.respect_robots_txt);
+        debug.field("soft_http_errors", &self.soft_http_errors);
+        debug.field("user_agent", &self.user_agent);
+        debug.field("stay_on_domain", &self.stay_on_domain);
+        debug.field("allow_subdomains", &self.allow_subdomains);
+        debug.field("include_paths", &self.include_paths);
+        debug.field("exclude_paths", &self.exclude_paths);
+        debug.field("path_patterns_match_query", &self.path_patterns_match_query);
+        debug.field("path_patterns_match_url", &self.path_patterns_match_url);
+        debug.field("dedup_include_query", &self.dedup_include_query);
+        debug.field("strip_tracking_params", &self.strip_tracking_params);
+        debug.field("tracking_params", &self.tracking_params);
+        debug.field("custom_headers", &"<redacted>");
+        debug.field("request_timeout", &self.request_timeout);
+        debug.field("rate_limit_ms", &self.rate_limit_ms);
+        debug.field("max_redirects", &self.max_redirects);
+        debug.field("retry_count", &self.retry_count);
+        debug.field("retry_codes", &self.retry_codes);
+        debug.field("retry_initial_delay_ms", &self.retry_initial_delay_ms);
+        debug.field("retry_max_delay_ms", &self.retry_max_delay_ms);
+        debug.field("rate_limit_jitter_ratio", &self.rate_limit_jitter_ratio);
+        debug.field("cookies_enabled", &self.cookies_enabled);
+        debug.field("auth", &self.auth);
+        debug.field("max_body_size", &self.max_body_size);
+        debug.field("remove_tags", &self.remove_tags);
+        debug.field("content", &self.content);
+        debug.field("map_limit", &self.map_limit);
+        debug.field("map_search", &self.map_search);
+        debug.field("download_assets", &self.download_assets);
+        debug.field("asset_types", &self.asset_types);
+        debug.field("max_asset_size", &self.max_asset_size);
+        debug.field("browser", &self.browser);
+        debug.field("proxy", &self.proxy);
+        debug.field("user_agents", &self.user_agents);
+        debug.field("capture_screenshot", &self.capture_screenshot);
+        debug.field("follow_document_urls", &self.follow_document_urls);
+        debug.field("document_url_depth", &self.document_url_depth);
+        debug.field("download_documents", &self.download_documents);
+        debug.field("document_max_size", &self.document_max_size);
+        debug.field("document_mime_types", &self.document_mime_types);
+        debug.field("document_output_dir", &self.document_output_dir);
+        debug.field("document_content_encoding", &self.document_content_encoding);
+        debug.field("warc_output", &self.warc_output);
+        debug.field("browser_profile", &self.browser_profile);
+        debug.field("save_browser_profile", &self.save_browser_profile);
+        debug.field("ssrf", &self.ssrf);
+        debug.field("ssrf_deny_private_explicit", &self.ssrf_deny_private_explicit);
+        debug.finish()
+    }
+}
 
 impl Default for CrawlConfig {
     fn default() -> Self {
@@ -5241,6 +5368,15 @@ impl CrawlConfig {
                 })?,
                 None => false,
             },
+            path_patterns_match_url: match kwargs.get(ruby.to_symbol("path_patterns_match_url")) {
+                Some(v) => bool::try_convert(v).map_err(|e| {
+                    magnus::Error::new(
+                        unsafe { magnus::Ruby::get_unchecked() }.exception_type_error(),
+                        format!("invalid value for `path_patterns_match_url`: {}", e),
+                    )
+                })?,
+                None => false,
+            },
             dedup_include_query: match kwargs.get(ruby.to_symbol("dedup_include_query")) {
                 Some(v) => bool::try_convert(v).map_err(|e| {
                     magnus::Error::new(
@@ -5664,6 +5800,10 @@ impl CrawlConfig {
         self.path_patterns_match_query
     }
 
+    fn path_patterns_match_url(&self) -> bool {
+        self.path_patterns_match_url
+    }
+
     fn dedup_include_query(&self) -> bool {
         self.dedup_include_query
     }
@@ -29315,7 +29455,7 @@ impl PropertyChange {
     }
 }
 
-#[derive(Clone, Debug, serde::Serialize, serde::Deserialize)]
+#[derive(Clone, serde::Serialize, serde::Deserialize)]
 #[serde(default)]
 #[magnus::wrap(class = "Xberg::ProxyConfig")]
 pub struct ProxyConfig {
@@ -29351,6 +29491,15 @@ impl magnus::TryConvert for ProxyConfig {
 }
 
 unsafe impl TryConvertOwned for ProxyConfig {}
+impl std::fmt::Debug for ProxyConfig {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        let mut debug = f.debug_struct("ProxyConfig");
+        debug.field("url", &"<redacted>");
+        debug.field("username", &self.username);
+        debug.field("password", &"<redacted>");
+        debug.finish()
+    }
+}
 
 impl Default for ProxyConfig {
     fn default() -> Self {
@@ -32038,6 +32187,7 @@ impl SparseEmbeddingPreset {
 pub struct SsrfPolicy {
     deny_private: bool,
     allowlist: Vec<HostMatcher>,
+    denylist: Vec<HostMatcher>,
     max_redirects: u8,
     scheme_allowlist: Vec<String>,
 }
@@ -32109,6 +32259,15 @@ impl SsrfPolicy {
                 })?,
                 None => Default::default(),
             },
+            denylist: match kwargs.get(ruby.to_symbol("denylist")) {
+                Some(v) => <Vec<HostMatcher>>::try_convert(v).map_err(|e| {
+                    magnus::Error::new(
+                        unsafe { magnus::Ruby::get_unchecked() }.exception_type_error(),
+                        format!("invalid value for `denylist`: {}", e),
+                    )
+                })?,
+                None => Default::default(),
+            },
             max_redirects: match kwargs.get(ruby.to_symbol("max_redirects")) {
                 Some(v) => u8::try_convert(v).map_err(|e| {
                     magnus::Error::new(
@@ -32138,6 +32297,10 @@ impl SsrfPolicy {
         self.allowlist.clone()
     }
 
+    fn denylist(&self) -> Vec<HostMatcher> {
+        self.denylist.clone()
+    }
+
     fn max_redirects(&self) -> u8 {
         self.max_redirects
     }
@@ -36266,7 +36429,7 @@ impl magnus::TryConvert for AssetCategory {
 unsafe impl IntoValueFromNative for AssetCategory {}
 unsafe impl TryConvertOwned for AssetCategory {}
 
-#[derive(Clone, Debug, serde::Serialize, serde::Deserialize)]
+#[derive(Clone, serde::Serialize, serde::Deserialize)]
 #[serde(tag = "type")]
 pub enum AuthConfig {
     #[serde(rename = "basic")]
@@ -36332,6 +36495,11 @@ unsafe impl IntoValueFromNative for AuthConfig {}
 unsafe impl TryConvertOwned for AuthConfig {}
 
 impl AuthConfig {}
+impl std::fmt::Debug for AuthConfig {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        f.write_str("AuthConfig(<redacted>)")
+    }
+}
 
 #[derive(Clone, Copy, PartialEq, Eq, Hash, Debug, serde::Serialize, serde::Deserialize)]
 #[serde(rename_all = "snake_case")]
@@ -48325,6 +48493,8 @@ impl From<BrowserConfig> for crawlberg::BrowserConfig {
             robots_user_agent: val.robots_user_agent,
             capture_network_events: val.capture_network_events,
             session_affinity: val.session_affinity,
+            chrome_path: val.chrome_path.map(Into::into),
+            chrome_args: val.chrome_args.into_iter().collect(),
             ..Default::default()
         }
     }
@@ -48349,6 +48519,8 @@ impl From<crawlberg::BrowserConfig> for BrowserConfig {
             robots_user_agent: val.robots_user_agent.map(|v| v.to_string()),
             capture_network_events: val.capture_network_events,
             session_affinity: val.session_affinity,
+            chrome_path: val.chrome_path.map(|p| p.to_string_lossy().to_string()),
+            chrome_args: val.chrome_args.into_iter().collect(),
         }
     }
 }
@@ -48918,6 +49090,7 @@ impl From<ConversionOptions> for html_to_markdown_rs::ConversionOptions {
             include_document_structure: val.include_document_structure,
             extract_images: val.extract_images,
             max_image_size: val.max_image_size,
+            max_input_size: val.max_input_size,
             capture_svg: val.capture_svg,
             infer_dimensions: val.infer_dimensions,
             max_depth: val.max_depth,
@@ -48972,6 +49145,7 @@ impl From<html_to_markdown_rs::ConversionOptions> for ConversionOptions {
             include_document_structure: val.include_document_structure,
             extract_images: val.extract_images,
             max_image_size: val.max_image_size,
+            max_input_size: val.max_input_size,
             capture_svg: val.capture_svg,
             infer_dimensions: val.infer_dimensions,
             max_depth: val.max_depth,
@@ -49052,6 +49226,7 @@ impl From<CrawlConfig> for crawlberg::CrawlConfig {
             include_paths: val.include_paths.into_iter().collect(),
             exclude_paths: val.exclude_paths.into_iter().collect(),
             path_patterns_match_query: val.path_patterns_match_query,
+            path_patterns_match_url: val.path_patterns_match_url,
             dedup_include_query: val.dedup_include_query,
             strip_tracking_params: val.strip_tracking_params,
             tracking_params: val.tracking_params.into_iter().collect(),
@@ -49119,6 +49294,7 @@ impl From<crawlberg::CrawlConfig> for CrawlConfig {
             include_paths: val.include_paths.into_iter().collect(),
             exclude_paths: val.exclude_paths.into_iter().collect(),
             path_patterns_match_query: val.path_patterns_match_query,
+            path_patterns_match_url: val.path_patterns_match_url,
             dedup_include_query: val.dedup_include_query,
             strip_tracking_params: val.strip_tracking_params,
             tracking_params: val.tracking_params.into_iter().collect(),
@@ -53939,6 +54115,7 @@ impl From<SsrfPolicy> for crawlberg::SsrfPolicy {
         Self {
             deny_private: val.deny_private,
             allowlist: val.allowlist.into_iter().map(Into::into).collect(),
+            denylist: val.denylist.into_iter().map(Into::into).collect(),
             max_redirects: val.max_redirects,
             scheme_allowlist: val.scheme_allowlist.into_iter().collect(),
             ..Default::default()
@@ -53952,6 +54129,7 @@ impl From<crawlberg::SsrfPolicy> for SsrfPolicy {
         Self {
             deny_private: val.deny_private,
             allowlist: val.allowlist.into_iter().map(Into::into).collect(),
+            denylist: val.denylist.into_iter().map(Into::into).collect(),
             max_redirects: val.max_redirects,
             scheme_allowlist: val.scheme_allowlist.into_iter().collect(),
         }
@@ -57397,30 +57575,30 @@ impl From<html_to_markdown_rs::WhitespaceMode> for WhitespaceMode {
     }
 }
 
-/// Convert a `xberg::heuristics::error::HeuristicsError` error to a Magnus runtime error.
+/// Convert a `xberg::HeuristicsError` error to a Magnus runtime error.
 #[allow(dead_code)]
-fn heuristics_error_to_magnus_err(e: xberg::heuristics::error::HeuristicsError) -> magnus::Error {
+fn heuristics_error_to_magnus_err(e: xberg::HeuristicsError) -> magnus::Error {
     let msg = e.to_string();
     magnus::Error::new(unsafe { magnus::Ruby::get_unchecked() }.exception_runtime_error(), msg)
 }
 
-/// Convert a `xberg::presets::loader::LoadError` error to a Magnus runtime error.
+/// Convert a `xberg::LoadError` error to a Magnus runtime error.
 #[allow(dead_code)]
-fn load_error_to_magnus_err(e: xberg::presets::loader::LoadError) -> magnus::Error {
+fn load_error_to_magnus_err(e: xberg::LoadError) -> magnus::Error {
     let msg = e.to_string();
     magnus::Error::new(unsafe { magnus::Ruby::get_unchecked() }.exception_runtime_error(), msg)
 }
 
-/// Convert a `xberg::presets::resolve::ResolveError` error to a Magnus runtime error.
+/// Convert a `xberg::ResolveError` error to a Magnus runtime error.
 #[allow(dead_code)]
-fn resolve_error_to_magnus_err(e: xberg::presets::resolve::ResolveError) -> magnus::Error {
+fn resolve_error_to_magnus_err(e: xberg::ResolveError) -> magnus::Error {
     let msg = e.to_string();
     magnus::Error::new(unsafe { magnus::Ruby::get_unchecked() }.exception_runtime_error(), msg)
 }
 
-/// Convert a `xberg::error::XbergError` error to a Magnus runtime error.
+/// Convert a `xberg::XbergError` error to a Magnus runtime error.
 #[allow(dead_code)]
-fn xberg_error_to_magnus_err(e: xberg::error::XbergError) -> magnus::Error {
+fn xberg_error_to_magnus_err(e: xberg::XbergError) -> magnus::Error {
     let msg = e.to_string();
     magnus::Error::new(unsafe { magnus::Ruby::get_unchecked() }.exception_runtime_error(), msg)
 }
@@ -57591,6 +57769,10 @@ fn ruby_init(ruby: &Ruby) -> Result<(), Error> {
 
     class.define_method("session_affinity", method!(BrowserConfig::session_affinity, 0))?;
 
+    class.define_method("chrome_path", method!(BrowserConfig::chrome_path, 0))?;
+
+    class.define_method("chrome_args", method!(BrowserConfig::chrome_args, 0))?;
+
     let class = module.define_class("CacheStats", ruby.class_object())?;
 
     class.define_singleton_method("new", function!(CacheStats::new, -1))?;
@@ -58027,6 +58209,8 @@ fn ruby_init(ruby: &Ruby) -> Result<(), Error> {
 
     class.define_method("max_image_size", method!(ConversionOptions::max_image_size, 0))?;
 
+    class.define_method("max_input_size", method!(ConversionOptions::max_input_size, 0))?;
+
     class.define_method("capture_svg", method!(ConversionOptions::capture_svg, 0))?;
 
     class.define_method("infer_dimensions", method!(ConversionOptions::infer_dimensions, 0))?;
@@ -58128,6 +58312,11 @@ fn ruby_init(ruby: &Ruby) -> Result<(), Error> {
         method!(CrawlConfig::path_patterns_match_query, 0),
     )?;
 
+    class.define_method(
+        "path_patterns_match_url",
+        method!(CrawlConfig::path_patterns_match_url, 0),
+    )?;
+
     class.define_method("dedup_include_query", method!(CrawlConfig::dedup_include_query, 0))?;
 
     class.define_method("strip_tracking_params", method!(CrawlConfig::strip_tracking_params, 0))?;
@@ -61805,6 +61994,8 @@ fn ruby_init(ruby: &Ruby) -> Result<(), Error> {
 
     class.define_method("allowlist", method!(SsrfPolicy::allowlist, 0))?;
 
+    class.define_method("denylist", method!(SsrfPolicy::denylist, 0))?;
+
     class.define_method("max_redirects", method!(SsrfPolicy::max_redirects, 0))?;
 
     class.define_method("scheme_allowlist", method!(SsrfPolicy::scheme_allowlist, 0))?;

data/lib/xberg/version.rb

@@ -1,10 +1,10 @@
 # This file is auto-generated by alef — DO NOT EDIT.
-# alef:hash:a73b2b7b23be41b809d7a63c6e84f6b22dc213b4d2d018e2004ef2a2f0e9410e
+# alef:hash:d17ea85ac3bb662944c743ac8f433a874d8fb19b8ce0dc8bc7e899d6d329f650
 # To regenerate: alef generate
 # To verify freshness: alef verify
 # frozen_string_literal: true
 
 module Xberg
   ## The version string for this package.
-  VERSION = "1.3.4"
+  VERSION = "1.3.5"
 end

data/sig/types.rbs

@@ -1,5 +1,5 @@
 # This file is auto-generated by alef — DO NOT EDIT.
-# alef:hash:4d6e82c5b0834d91c2575ca21fe65ed7dc5bb424ab80623cb550f60e8cd268ad
+# alef:hash:d01bd27d5fa9368a18dda14ba966abab79e6f319668303eaa72b202e75e41946
 # To regenerate: alef generate
 # To verify freshness: alef verify
 
@@ -107,8 +107,10 @@ def initialize: (?entry_count: Integer, ?citation_keys: Array[String], ?authors:
         attr_reader robots_user_agent: String?
         attr_reader capture_network_events: bool
         attr_reader session_affinity: bool
+        attr_reader chrome_path: String?
+    attr_reader chrome_args: Array[String]
 
-def initialize: (?mode: enum_BrowserMode, ?backend: enum_BrowserBackend, ?endpoint: String, ?timeout: Integer, ?overall_timeout: Integer, ?shutdown_timeout: Integer, ?wait: enum_BrowserWait, ?wait_selector: String, ?extra_wait: Integer, ?proxy: ProxyConfig, ?block_url_patterns: Array[String], ?eval_script: String, ?robots_user_agent: String, ?capture_network_events: bool, ?session_affinity: bool) -> void
+def initialize: (?mode: enum_BrowserMode, ?backend: enum_BrowserBackend, ?endpoint: String, ?timeout: Integer, ?overall_timeout: Integer, ?shutdown_timeout: Integer, ?wait: enum_BrowserWait, ?wait_selector: String, ?extra_wait: Integer, ?proxy: ProxyConfig, ?block_url_patterns: Array[String], ?eval_script: String, ?robots_user_agent: String, ?capture_network_events: bool, ?session_affinity: bool, ?chrome_path: String, ?chrome_args: Array[String]) -> void
     end
 
     class CacheStats
@@ -362,6 +364,7 @@ def initialize: (?output_format: String, ?preprocessing_preset: String, ?remove_
         attr_reader include_document_structure: bool
         attr_reader extract_images: bool
         attr_reader max_image_size: Integer
+        attr_reader max_input_size: Integer?
         attr_reader capture_svg: bool
         attr_reader infer_dimensions: bool
         attr_reader max_depth: Integer?
@@ -369,7 +372,7 @@ def initialize: (?output_format: String, ?preprocessing_preset: String, ?remove_
         attr_reader tier_strategy: enum_TierStrategy
         attr_reader base_url: String?
 
-def initialize: (?heading_style: enum_HeadingStyle, ?list_indent_type: enum_ListIndentType, ?list_indent_width: Integer, ?bullets: String, ?strong_em_symbol: String, ?escape_asterisks: bool, ?escape_underscores: bool, ?escape_misc: bool, ?escape_ascii: bool, ?code_language: String, ?autolinks: bool, ?default_title: bool, ?br_in_tables: bool, ?compact_tables: bool, ?highlight_style: enum_HighlightStyle, ?extract_metadata: bool, ?whitespace_mode: enum_WhitespaceMode, ?strip_newlines: bool, ?wrap: bool, ?wrap_width: Integer, ?convert_as_inline: bool, ?sub_symbol: String, ?sup_symbol: String, ?newline_style: enum_NewlineStyle, ?code_block_style: enum_CodeBlockStyle, ?keep_inline_images_in: Array[String], ?preprocessing: PreprocessingOptions, ?encoding: String, ?debug: bool, ?strip_tags: Array[String], ?preserve_tags: Array[String], ?skip_images: bool, ?inline_data_media: enum_InlineDataMedia, ?url_escape_style: enum_UrlEscapeStyle, ?link_style: enum_LinkStyle, ?output_format: enum_HtmlToMarkdownRsOutputFormat, ?include_document_structure: bool, ?extract_images: bool, ?max_image_size: Integer, ?capture_svg: bool, ?infer_dimensions: bool, ?max_depth: Integer, ?exclude_selectors: Array[String], ?tier_strategy: enum_TierStrategy, ?base_url: String) -> void
+def initialize: (?heading_style: enum_HeadingStyle, ?list_indent_type: enum_ListIndentType, ?list_indent_width: Integer, ?bullets: String, ?strong_em_symbol: String, ?escape_asterisks: bool, ?escape_underscores: bool, ?escape_misc: bool, ?escape_ascii: bool, ?code_language: String, ?autolinks: bool, ?default_title: bool, ?br_in_tables: bool, ?compact_tables: bool, ?highlight_style: enum_HighlightStyle, ?extract_metadata: bool, ?whitespace_mode: enum_WhitespaceMode, ?strip_newlines: bool, ?wrap: bool, ?wrap_width: Integer, ?convert_as_inline: bool, ?sub_symbol: String, ?sup_symbol: String, ?newline_style: enum_NewlineStyle, ?code_block_style: enum_CodeBlockStyle, ?keep_inline_images_in: Array[String], ?preprocessing: PreprocessingOptions, ?encoding: String, ?debug: bool, ?strip_tags: Array[String], ?preserve_tags: Array[String], ?skip_images: bool, ?inline_data_media: enum_InlineDataMedia, ?url_escape_style: enum_UrlEscapeStyle, ?link_style: enum_LinkStyle, ?output_format: enum_HtmlToMarkdownRsOutputFormat, ?include_document_structure: bool, ?extract_images: bool, ?max_image_size: Integer, ?max_input_size: Integer, ?capture_svg: bool, ?infer_dimensions: bool, ?max_depth: Integer, ?exclude_selectors: Array[String], ?tier_strategy: enum_TierStrategy, ?base_url: String) -> void
     end
 
     class CoreProperties
@@ -409,6 +412,7 @@ def initialize: (?heading_style: enum_HeadingStyle, ?list_indent_type: enum_List
     attr_reader include_paths: Array[String]
     attr_reader exclude_paths: Array[String]
         attr_reader path_patterns_match_query: bool
+        attr_reader path_patterns_match_url: bool
         attr_reader dedup_include_query: bool
         attr_reader strip_tracking_params: bool
     attr_reader tracking_params: Array[String]
@@ -448,7 +452,7 @@ def initialize: (?heading_style: enum_HeadingStyle, ?list_indent_type: enum_List
         attr_reader ssrf: SsrfPolicy
         attr_reader ssrf_deny_private_explicit: bool?
 
-def initialize: (?max_depth: Integer, ?max_pages: Integer, ?max_links_per_page: Integer, ?max_concurrent: Integer, ?crawl_strategy: enum_CrawlStrategyKind, ?content_filter: enum_ContentFilterKind, ?bm25_query: String, ?bm25_threshold: Float, ?respect_robots_txt: bool, ?soft_http_errors: bool, ?user_agent: String, ?stay_on_domain: bool, ?allow_subdomains: bool, ?include_paths: Array[String], ?exclude_paths: Array[String], ?path_patterns_match_query: bool, ?dedup_include_query: bool, ?strip_tracking_params: bool, ?tracking_params: Array[String], ?custom_headers: Hash[String, String], ?request_timeout: Integer, ?rate_limit_ms: Integer, ?max_redirects: Integer, ?retry_count: Integer, ?retry_codes: Array[Integer], ?retry_initial_delay_ms: Integer, ?retry_max_delay_ms: Integer, ?rate_limit_jitter_ratio: Float, ?cookies_enabled: bool, ?auth: AuthConfig, ?max_body_size: Integer, ?remove_tags: Array[String], ?content: ContentConfig, ?map_limit: Integer, ?map_search: String, ?download_assets: bool, ?asset_types: Array[enum_AssetCategory], ?max_asset_size: Integer, ?browser: BrowserConfig, ?proxy: ProxyConfig, ?user_agents: Array[String], ?capture_screenshot: bool, ?follow_document_urls: bool, ?document_url_depth: Integer, ?download_documents: bool, ?document_max_size: Integer, ?document_mime_types: Array[String], ?document_output_dir: String, ?document_content_encoding: enum_DocumentContentEncoding, ?warc_output: String, ?browser_profile: String, ?save_browser_profile: bool, ?ssrf: SsrfPolicy, ?ssrf_deny_private_explicit: bool) -> void
+def initialize: (?max_depth: Integer, ?max_pages: Integer, ?max_links_per_page: Integer, ?max_concurrent: Integer, ?crawl_strategy: enum_CrawlStrategyKind, ?content_filter: enum_ContentFilterKind, ?bm25_query: String, ?bm25_threshold: Float, ?respect_robots_txt: bool, ?soft_http_errors: bool, ?user_agent: String, ?stay_on_domain: bool, ?allow_subdomains: bool, ?include_paths: Array[String], ?exclude_paths: Array[String], ?path_patterns_match_query: bool, ?path_patterns_match_url: bool, ?dedup_include_query: bool, ?strip_tracking_params: bool, ?tracking_params: Array[String], ?custom_headers: Hash[String, String], ?request_timeout: Integer, ?rate_limit_ms: Integer, ?max_redirects: Integer, ?retry_count: Integer, ?retry_codes: Array[Integer], ?retry_initial_delay_ms: Integer, ?retry_max_delay_ms: Integer, ?rate_limit_jitter_ratio: Float, ?cookies_enabled: bool, ?auth: AuthConfig, ?max_body_size: Integer, ?remove_tags: Array[String], ?content: ContentConfig, ?map_limit: Integer, ?map_search: String, ?download_assets: bool, ?asset_types: Array[enum_AssetCategory], ?max_asset_size: Integer, ?browser: BrowserConfig, ?proxy: ProxyConfig, ?user_agents: Array[String], ?capture_screenshot: bool, ?follow_document_urls: bool, ?document_url_depth: Integer, ?download_documents: bool, ?document_max_size: Integer, ?document_mime_types: Array[String], ?document_output_dir: String, ?document_content_encoding: enum_DocumentContentEncoding, ?warc_output: String, ?browser_profile: String, ?save_browser_profile: bool, ?ssrf: SsrfPolicy, ?ssrf_deny_private_explicit: bool) -> void
     end
 
     class CsvConfig
@@ -2444,10 +2448,11 @@ def initialize: (name: String, model_repo: String, model_file: String, additiona
     class SsrfPolicy
         attr_reader deny_private: bool
     attr_reader allowlist: Array[HostMatcher]
+    attr_reader denylist: Array[HostMatcher]
         attr_reader max_redirects: Integer
     attr_reader scheme_allowlist: Array[String]
 
-def initialize: (?deny_private: bool, ?allowlist: Array[HostMatcher], ?max_redirects: Integer, ?scheme_allowlist: Array[String]) -> void
+def initialize: (?deny_private: bool, ?allowlist: Array[HostMatcher], ?denylist: Array[HostMatcher], ?max_redirects: Integer, ?scheme_allowlist: Array[String]) -> void
     end
 
     class StructuredData

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: xberg
 version: !ruby/object:Gem::Version
-  version: 1.3.4
+  version: 1.3.5
 platform: ruby
 authors:
 - Na'aman Hirschfeld <naaman@xberg.io>
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2026-10-05 00:00:00.000000000 Z
+date: 2026-10-06 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rb_sys