Compare commits

..

1 Commits

Author SHA1 Message Date
1b9964a938 not yet 2025-03-20 14:33:27 -06:00
19 changed files with 844 additions and 917 deletions

4
.gitignore vendored
View File

@ -4,6 +4,4 @@
perf.data
flamegraph.svg
perf.data.old
/docker/logs/*
/downloaded
/Crawler.toml
/docker/logs/*

19
.vscode/launch.json vendored
View File

@ -7,15 +7,18 @@
{
"type": "lldb",
"request": "launch",
"name": "Debug executable 'internet_mapper'",
"name": "Debug executable 'surreal_spider'",
"env": {
"RUST_LOG": "surreal_spider=trace,reqwest=info",
},
"cargo": {
"args": [
"build",
"--bin=internet_mapper",
"--package=internet_mapper"
"--bin=surreal_spider",
"--package=surreal_spider"
],
"filter": {
"name": "internet_mapper",
"name": "surreal_spider",
"kind": "bin"
}
},
@ -25,16 +28,16 @@
{
"type": "lldb",
"request": "launch",
"name": "Debug unit tests in executable 'internet_mapper'",
"name": "Debug unit tests in executable 'surreal_spider'",
"cargo": {
"args": [
"test",
"--no-run",
"--bin=internet_mapper",
"--package=internet_mapper"
"--bin=surreal_spider",
"--package=surreal_spider"
],
"filter": {
"name": "internet_mapper",
"name": "surreal_spider",
"kind": "bin"
}
},

View File

@ -3,6 +3,6 @@
"creds",
"reqwest",
"rustls",
"surql",
"surql"
]
}

430
Cargo.lock generated
View File

@ -103,18 +103,62 @@ dependencies = [
"libc",
]
[[package]]
name = "anstream"
version = "0.6.18"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8acc5369981196006228e28809f761875c0327210a891e941f4c683b3a99529b"
dependencies = [
"anstyle",
"anstyle-parse",
"anstyle-query",
"anstyle-wincon",
"colorchoice",
"is_terminal_polyfill",
"utf8parse",
]
[[package]]
name = "anstyle"
version = "1.0.10"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "55cc3b69f167a1ef2e161439aa98aed94e6028e5f9a59be9a6ffb47aef1651f9"
[[package]]
name = "anstyle-parse"
version = "0.2.6"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3b2d16507662817a6a20a9ea92df6652ee4f94f914589377d69f3b21bc5798a9"
dependencies = [
"utf8parse",
]
[[package]]
name = "anstyle-query"
version = "1.1.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "79947af37f4177cfead1110013d678905c37501914fba0efea834c3fe9a8d60c"
dependencies = [
"windows-sys 0.59.0",
]
[[package]]
name = "anstyle-wincon"
version = "3.0.7"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ca3534e77181a9cc07539ad51f2141fe32f6c3ffd4df76db8ad92346b003ae4e"
dependencies = [
"anstyle",
"once_cell",
"windows-sys 0.59.0",
]
[[package]]
name = "any_ascii"
version = "0.3.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ea50b14b7a4b9343f8c627a7a53c52076482bd4bdad0a24fd3ec533ed616cc2c"
[[package]]
name = "anyhow"
version = "1.0.98"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e16d2d3311acee920a9eb8d33b8cbc1787ce4a264e85f964c2404b969bdcd487"
[[package]]
name = "approx"
version = "0.4.0"
@ -281,6 +325,17 @@ dependencies = [
"serde_json",
]
[[package]]
name = "async-recursion"
version = "1.1.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3b43422f69d8ff38f95f1b2bb76517c91589a924d1559a0e935d7c8ce0274c11"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
]
[[package]]
name = "async-stream"
version = "0.3.6"
@ -760,6 +815,12 @@ dependencies = [
"cc",
]
[[package]]
name = "colorchoice"
version = "1.0.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "5b63caa9aa9397e2d9480a9b13673856c78d8ac123288526c37d7839f2a86990"
[[package]]
name = "concurrent-queue"
version = "2.5.0"
@ -810,6 +871,21 @@ dependencies = [
"libc",
]
[[package]]
name = "crc"
version = "3.2.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "69e6e4d7b33a94f0991c26729976b10ebde1d34c3ee82408fb536164fa10d636"
dependencies = [
"crc-catalog",
]
[[package]]
name = "crc-catalog"
version = "2.4.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "19d374276b40fb8bbdee95aef7c7fa6b5316ec764510eb64b8dd0e2ed0d7e7f5"
[[package]]
name = "crc32fast"
version = "1.4.2"
@ -908,6 +984,20 @@ dependencies = [
"parking_lot_core",
]
[[package]]
name = "dashmap"
version = "6.1.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "5041cc499144891f3790297212f32a74fb938e5136a14943f338ef9e0ae276cf"
dependencies = [
"cfg-if",
"crossbeam-utils",
"hashbrown 0.14.5",
"lock_api",
"once_cell",
"parking_lot_core",
]
[[package]]
name = "data-encoding"
version = "2.8.0"
@ -924,6 +1014,17 @@ dependencies = [
"serde",
]
[[package]]
name = "derivative"
version = "2.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "fcc3dd5e9e9c0b295d6e1e4d811fb6f157d5ffd784b8d202fc62eac8035a770b"
dependencies = [
"proc-macro2",
"quote",
"syn 1.0.109",
]
[[package]]
name = "deunicode"
version = "1.6.1"
@ -1034,6 +1135,29 @@ version = "0.1.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "c34f04666d835ff5d62e058c3995147c06f42fe86ff053337632bca83e42702d"
[[package]]
name = "env_filter"
version = "0.1.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "186e05a59d4c50738528153b83b0b0194d3a29507dfec16eccd4b342903397d0"
dependencies = [
"log",
"regex",
]
[[package]]
name = "env_logger"
version = "0.11.7"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "c3716d7a920fb4fac5d84e9d4bce8ceb321e9414b4409da61b07b75c1e3d0697"
dependencies = [
"anstream",
"anstyle",
"env_filter",
"jiff",
"log",
]
[[package]]
name = "equivalent"
version = "1.0.2"
@ -1603,19 +1727,6 @@ dependencies = [
"webpki-roots",
]
[[package]]
name = "hyper-timeout"
version = "0.5.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2b90d566bffbce6a75bd8b09a05aa8c2cb1fabb6cb348f8840c9e4c90a0d83b0"
dependencies = [
"hyper",
"hyper-util",
"pin-project-lite",
"tokio",
"tower-service",
]
[[package]]
name = "hyper-tls"
version = "0.6.0"
@ -1855,14 +1966,10 @@ name = "internet_mapper"
version = "0.1.0"
dependencies = [
"base64 0.22.1",
"futures-util",
"html5ever 0.29.1",
"metrics",
"metrics-exporter-prometheus",
"opentelemetry",
"opentelemetry-otlp",
"opentelemetry_sdk",
"rand 0.9.1",
"minio",
"reqwest",
"serde",
"surrealdb",
@ -1879,6 +1986,12 @@ version = "2.11.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "469fb0b9cefa57e3ef31275ee7cacb78f2fdca44e4765491884a2b119d4eb130"
[[package]]
name = "is_terminal_polyfill"
version = "1.70.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "7943c866cc5cd64cbc25b2e01621d07fa8eb2a1a23160ee81ce38704e97b8ecf"
[[package]]
name = "itertools"
version = "0.10.5"
@ -1912,6 +2025,30 @@ version = "1.0.15"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "4a5f13b858c8d314ee3e8f639011f7ccefe71f97f96e50151fb991f267928e2c"
[[package]]
name = "jiff"
version = "0.2.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d699bc6dfc879fb1bf9bdff0d4c56f0884fc6f0d0eb0fba397a6d00cd9a6b85e"
dependencies = [
"jiff-static",
"log",
"portable-atomic",
"portable-atomic-util",
"serde",
]
[[package]]
name = "jiff-static"
version = "0.2.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8d16e75759ee0aa64c57a56acbf43916987b20c77373cb7e808979e02b93c9f9"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
]
[[package]]
name = "jobserver"
version = "0.1.32"
@ -2156,6 +2293,12 @@ dependencies = [
"digest",
]
[[package]]
name = "md5"
version = "0.7.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "490cc448043f947bae3cbee9c203358d62dbee0db12107a74be5c30ccfd09771"
[[package]]
name = "memchr"
version = "2.7.4"
@ -2254,6 +2397,46 @@ version = "0.2.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "68354c5c6bd36d73ff3feceb05efa59b6acb7626617f4962be322a825e61f79a"
[[package]]
name = "minio"
version = "0.2.0-alpha"
source = "git+https://github.com/minio/minio-rs.git?rev=c28f576#c28f576cb8f8cf47fb941bb9db62b2cbd6f080c1"
dependencies = [
"async-recursion",
"async-trait",
"base64 0.22.1",
"byteorder",
"bytes",
"chrono",
"crc",
"dashmap 6.1.0",
"derivative",
"env_logger",
"futures-util",
"hex",
"hmac",
"home",
"http",
"hyper",
"lazy_static",
"log",
"md5",
"multimap",
"os_info",
"percent-encoding",
"rand 0.8.5",
"regex",
"reqwest",
"serde",
"serde_json",
"sha2",
"tokio",
"tokio-stream",
"tokio-util",
"urlencoding",
"xmltree",
]
[[package]]
name = "miniz_oxide"
version = "0.8.5"
@ -2291,6 +2474,15 @@ dependencies = [
"version_check",
]
[[package]]
name = "multimap"
version = "0.10.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "defc4c55412d89136f966bbb339008b474350e5e6e78d2714439c386b3137a03"
dependencies = [
"serde",
]
[[package]]
name = "nanoid"
version = "0.4.0"
@ -2543,77 +2735,14 @@ dependencies = [
]
[[package]]
name = "opentelemetry"
version = "0.30.0"
name = "os_info"
version = "3.10.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "aaf416e4cb72756655126f7dd7bb0af49c674f4c1b9903e80c009e0c37e552e6"
checksum = "2a604e53c24761286860eba4e2c8b23a0161526476b1de520139d69cdb85a6b5"
dependencies = [
"futures-core",
"futures-sink",
"js-sys",
"pin-project-lite",
"thiserror 2.0.12",
"tracing",
]
[[package]]
name = "opentelemetry-http"
version = "0.30.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "50f6639e842a97dbea8886e3439710ae463120091e2e064518ba8e716e6ac36d"
dependencies = [
"async-trait",
"bytes",
"http",
"opentelemetry",
"reqwest",
]
[[package]]
name = "opentelemetry-otlp"
version = "0.30.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "dbee664a43e07615731afc539ca60c6d9f1a9425e25ca09c57bc36c87c55852b"
dependencies = [
"http",
"opentelemetry",
"opentelemetry-http",
"opentelemetry-proto",
"opentelemetry_sdk",
"prost",
"reqwest",
"thiserror 2.0.12",
"tokio",
"tonic",
"tracing",
]
[[package]]
name = "opentelemetry-proto"
version = "0.30.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2e046fd7660710fe5a05e8748e70d9058dc15c94ba914e7c4faa7c728f0e8ddc"
dependencies = [
"opentelemetry",
"opentelemetry_sdk",
"prost",
"tonic",
]
[[package]]
name = "opentelemetry_sdk"
version = "0.30.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "11f644aa9e5e31d11896e024305d7e3c98a88884d9f8919dbf37a9991bc47a4b"
dependencies = [
"futures-channel",
"futures-executor",
"futures-util",
"opentelemetry",
"percent-encoding",
"rand 0.9.1",
"serde_json",
"thiserror 2.0.12",
"log",
"serde",
"windows-sys 0.52.0",
]
[[package]]
@ -2793,26 +2922,6 @@ version = "0.5.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "5be167a7af36ee22fe3115051bc51f6e6c7054c9348e28deb4f49bd6f705a315"
[[package]]
name = "pin-project"
version = "1.1.10"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "677f1add503faace112b9f1373e43e9e054bfdd22ff1a63c1bc485eaec6a6a8a"
dependencies = [
"pin-project-internal",
]
[[package]]
name = "pin-project-internal"
version = "1.1.10"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "6e918e4ff8c4549eb882f14b3a4bc8c8bc93de829416eacf579f1207a8fbf861"
dependencies = [
"proc-macro2",
"quote",
"syn 2.0.100",
]
[[package]]
name = "pin-project-lite"
version = "0.2.16"
@ -2837,6 +2946,15 @@ version = "1.11.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "350e9b48cbc6b0e028b0473b114454c6316e57336ee184ceab6e53f72c178b3e"
[[package]]
name = "portable-atomic-util"
version = "0.2.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d8a2f0d8d040d7848a709caf78912debcc3f33ee4b3cac47d73d1e1069e83507"
dependencies = [
"portable-atomic",
]
[[package]]
name = "powerfmt"
version = "0.2.0"
@ -2886,29 +3004,6 @@ dependencies = [
"unicode-ident",
]
[[package]]
name = "prost"
version = "0.13.5"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2796faa41db3ec313a31f7624d9286acf277b52de526150b7e69f3debf891ee5"
dependencies = [
"bytes",
"prost-derive",
]
[[package]]
name = "prost-derive"
version = "0.13.5"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8a56d757972c98b346a9b766e3f02746cde6dd1cd1d1d563472929fdd74bec4d"
dependencies = [
"anyhow",
"itertools 0.13.0",
"proc-macro2",
"quote",
"syn 2.0.100",
]
[[package]]
name = "psl-types"
version = "2.0.11"
@ -2999,7 +3094,7 @@ checksum = "b820744eb4dc9b57a3398183639c511b5a26d2ed702cedd3febaa1393caa22cc"
dependencies = [
"bytes",
"getrandom 0.3.2",
"rand 0.9.1",
"rand 0.9.0",
"ring",
"rustc-hash 2.1.1",
"rustls",
@ -3070,12 +3165,13 @@ dependencies = [
[[package]]
name = "rand"
version = "0.9.1"
version = "0.9.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "9fbfd9d094a40bf3ae768db9361049ace4c0e04a4fd6b359518bd7b73a73dd97"
checksum = "3779b94aeb87e8bd4e834cee3650289ee9e0d5677f976ecdb6d219e5f4f6cd94"
dependencies = [
"rand_chacha 0.9.0",
"rand_core 0.9.3",
"zerocopy 0.8.23",
]
[[package]]
@ -3269,7 +3365,6 @@ dependencies = [
"base64 0.22.1",
"bytes",
"encoding_rs",
"futures-channel",
"futures-core",
"futures-util",
"h2",
@ -4100,7 +4195,7 @@ dependencies = [
"cedar-policy",
"chrono",
"ciborium",
"dashmap",
"dashmap 5.5.3",
"deunicode",
"dmp",
"fst",
@ -4523,32 +4618,6 @@ dependencies = [
"winnow",
]
[[package]]
name = "tonic"
version = "0.13.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "7e581ba15a835f4d9ea06c55ab1bd4dce26fc53752c69a04aac00703bfb49ba9"
dependencies = [
"async-trait",
"base64 0.22.1",
"bytes",
"http",
"http-body",
"http-body-util",
"hyper",
"hyper-timeout",
"hyper-util",
"percent-encoding",
"pin-project",
"prost",
"tokio",
"tokio-stream",
"tower",
"tower-layer",
"tower-service",
"tracing",
]
[[package]]
name = "tower"
version = "0.5.2"
@ -4557,15 +4626,11 @@ checksum = "d039ad9159c98b70ecfd540b2573b97f7f52c3e8d9f8ad57a24b916a536975f9"
dependencies = [
"futures-core",
"futures-util",
"indexmap 2.8.0",
"pin-project-lite",
"slab",
"sync_wrapper",
"tokio",
"tokio-util",
"tower-layer",
"tower-service",
"tracing",
]
[[package]]
@ -4711,7 +4776,7 @@ version = "1.2.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "470dbf6591da1b39d43c14523b2b469c86879a53e8b758c8e090a470fe7b1fbe"
dependencies = [
"rand 0.9.1",
"rand 0.9.0",
"serde",
"web-time",
]
@ -4807,6 +4872,12 @@ version = "1.0.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b6c140620e7ffbb22c2dee59cafe6084a59b5ffc27a8859a5f0d494b5d52b6be"
[[package]]
name = "utf8parse"
version = "0.2.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "06abde3611657adf66d383f00b093d7faecc7fa57071cce2578660c9f1010821"
[[package]]
name = "uuid"
version = "1.16.0"
@ -5347,6 +5418,21 @@ dependencies = [
"tap",
]
[[package]]
name = "xml-rs"
version = "0.8.25"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "c5b940ebc25896e71dd073bad2dbaa2abfe97b0a391415e22ad1326d9c54e3c4"
[[package]]
name = "xmltree"
version = "0.11.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b619f8c85654798007fb10afa5125590b43b088c225a25fc2fec100a9fad0fc6"
dependencies = [
"xml-rs",
]
[[package]]
name = "yoke"
version = "0.7.5"

View File

@ -5,15 +5,12 @@ edition = "2021"
[dependencies]
base64 = "0.22.1"
futures-util = "0.3.31"
html5ever = "0.29"
metrics = "0.24.1"
metrics-exporter-prometheus = { version = "0.16.2", features=["http-listener"]}
opentelemetry = "0.30.0"
opentelemetry-otlp = { version = "0.30.0", features = ["metrics", "trace", "logs", "grpc-tonic"] }
opentelemetry_sdk = "0.30.0"
rand = "0.9.1"
reqwest = { version = "0.12", features = ["gzip", "default", "rustls-tls", "stream"] }
# minio = "0.1.0"
minio = {git="https://github.com/minio/minio-rs.git", rev = "c28f576"}
reqwest = { version = "0.12", features = ["gzip", "default", "rustls-tls"] }
serde = { version = "1.0", features = ["derive"] }
surrealdb = "2.2"
tokio = { version="1.41.0", features = ["full"] }

View File

@ -1,22 +1,16 @@
# Visability config
# Alloy (for Tempo)
tracing_endpoint = "http://localhost:4317"
# Prometheus
metrics_endpoint = "http://localhost:9090/api/v1/otlp/v1/metrics"
# Alloy (for Loki)
log_file = "./docker/logs/tracing.log"
# Surreal config
surreal_url = "localhost:8000"
surreal_username = "root"
surreal_password = "root"
surreal_ns = "test"
surreal_db = "v1.21.1"
surreal_db = "v1.16"
# Minio config
s3_bucket = "v1.16"
s3_url = "http://localhost:9000"
s3_access_key = "DwJfDDVIbmCmfAblwSqp"
s3_secret_key = "V4UqvC1Vm4AwLE5FAhu2gxlMfvexTBQnDxuy8uZx"
# Crawler config
crawl_filter = "https://ftpgeoinfo.msl.mt.gov/Data/Spatial/MSDI/Imagery/2023_NAIP/UTM_County_Mosaics/"
# crawl_filter = "https://oliveratkinson.net"
start_url = "https://ftpgeoinfo.msl.mt.gov/Data/Spatial/MSDI/Imagery/2023_NAIP/UTM_County_Mosaics/"
# start_url = "https://oliveratkinson.net"
budget = 100
batch_size = 2
crawl_filter = "en.wikipedia.com"
budget = 1000

View File

@ -2,54 +2,14 @@
Crawls sites saving all the found links to a surrealdb database. It then proceeds to take batches of 100 uncrawled links untill the crawl budget is reached. It saves the data of each site in a minio database.
## How to use
1. Clone the repo and `cd` into it.
2. Build the repo with `cargo build -r`
3. Start the docker conatiners
1. cd into the docker folder `cd docker`
2. Bring up the docker containers `docker compose up -d`
4. From the project's root, edit the `Crawler.toml` file to your liking.
5. Run with `./target/release/internet_mapper`
You can view stats of the project at `http://<your-ip>:3000/dashboards`
```bash
# Untested script but probably works
git clone https://git.oliveratkinson.net/Oliver/internet_mapper.git
cd internet_mapper
cargo build -r
cd docker
docker compose up -d
cd ..
$EDITOR Crawler.toml
./target/release/internet_mapper
```
### TODO
- [x] Domain filtering - prevent the crawler from going on alternate versions of wikipedia.
- [ ] Domain filtering - prevent the crawler from going on alternate versions of wikipedia.
- [ ] Conditionally save content - based on filename or file contents
- [x] GUI / TUI ? - Graphana
- [x] Better asynchronous getting of the sites. Currently it all happens serially.
- [x] Allow for storing asynchronously - dropping the "links to" logic fixes this need
- [x] Control crawler via config file (no recompliation needed)
3/17/25: Took >1hr to crawl 100 pages
3/19/25: Took 20min to crawl 1000 pages
This ment we stored 1000 pages, 142,997 urls, and 1,425,798 links between the two.
3/20/25: Took 5min to crawl 1000 pages
3/21/25: Took 3min to crawl 1000 pages
# About
![Screenshot](/pngs/graphana.png)
- [ ] Allow for storing asynchronously
3/19/25: Took 20min to crawl 100 pages
This ment we stored 100 pages, 142,997 urls, and 1,425,798 links between the two.

View File

@ -3,34 +3,12 @@ local.file_match "tmplogs" {
}
loki.source.file "local_files" {
targets = local.file_match.tmplogs.targets
forward_to = [loki.write.local_loki.receiver]
targets = local.file_match.tmplogs.targets
forward_to = [loki.write.local_loki.receiver]
}
loki.write "local_loki" {
endpoint {
url = "http://loki:3100/loki/api/v1/push"
}
}
otelcol.receiver.otlp "otlp_receiver" {
grpc {
endpoint = "0.0.0.0:4317"
}
http {
endpoint = "0.0.0.0:4318"
}
output {
traces = [otelcol.exporter.otlp.tempo.input,]
}
}
otelcol.exporter.otlp "tempo" {
client {
endpoint = "tempo:4317"
tls {
insecure = true
}
}
}
}

View File

@ -1,6 +1,4 @@
services:
# Database
surreal:
image: surrealdb/surrealdb:latest-dev
ports:
@ -16,19 +14,23 @@ services:
- --pass
- root
- rocksdb:/mydata/database.db
# Tracing
tempo:
image: grafana/tempo:latest
command: [ "-config.file=/etc/tempo.yaml" ]
volumes:
- ./tempo.yaml:/etc/tempo.yaml
- tempo_storage:/var/tempo
minio:
image: quay.io/minio/minio
ports:
- 3200:3200 # self metrics for prometheus
- 4317:4317 # otlp grpc - (alloy)
- 9000:9000
- 9001:9001
environment:
- MINIO_ROOT_USER=root
- MINIO_ROOT_PASSWORD=an8charpassword
- MINIO_PROMETHEUS_AUTH_TYPE=public
volumes:
- minio_storage:/data
command:
- server
- /data
- --console-address
- ":9001"
# Log scraper
alloy:
image: grafana/alloy:latest
ports:
@ -38,13 +40,9 @@ services:
- ./logs/:/tmp/alloy-logs
- ./alloy.conf:/etc/alloy/config.alloy
- alloy_storage:/var/lib/alloy
command:
- run
- --server.http.listen-addr=0.0.0.0:12345
- --storage.path=/var/lib/alloy/data
- /etc/alloy/config.alloy
command: run --server.http.listen-addr=0.0.0.0:12345 --storage.path=/var/lib/alloy/data /etc/alloy/config.alloy
# Log storage / analysis
#logs
loki:
image: grafana/loki:latest
ports:
@ -53,21 +51,16 @@ services:
volumes:
- ./loki.yaml:/etc/loki/local-config.yaml
# Metrics
# Metrics collector
prometheus:
image: prom/prometheus:latest
ports:
- 9090:9090
expose:
- 9090
volumes:
- ./prometheus.yaml:/etc/prometheus/prometheus.yml
# persist data
# - prometheus_storage:/prometheus
command:
- --enable-feature=native-histograms
- --web.enable-remote-write-receiver
- --web.enable-lifecycle
- --web.enable-otlp-receiver
- --config.file=/etc/prometheus/prometheus.yml
- prometheus_storage:/prometheus
command: --web.enable-lifecycle --config.file=/etc/prometheus/prometheus.yml
# Everything viewer
grafana:
@ -89,4 +82,4 @@ volumes:
grafana_storage:
alloy_storage:
surrealdb_storage:
tempo_storage:
minio_storage:

View File

@ -22,20 +22,3 @@ datasources:
editable: false
jsonData:
httpMethod: GET
- name: Tempo
type: tempo
access: proxy
orgId: 1
url: http://tempo:3200
basicAuth: false
isDefault: false
version: 1
editable: true
apiVersion: 1
uid: tempo
jsonData:
httpMethod: GET
serviceMap:
datasourceUid: prometheus
streamingEnabled:
search: true

View File

@ -1,15 +1,20 @@
global:
scrape_interval: 60s
scrape_interval: 5s
query_log_file: /etc/prometheus/query.log
scrape_configs:
# Crawler configs get pushed with OTLP
# - job_name: 'loki'
# static_configs:
# - targets: ['loki:3100']
# - job_name: 'prometheus'
# static_configs:
# - targets: ['localhost:9090']
- job_name: 'tempo'
- job_name: crawler
static_configs:
- targets: ['tempo:3200']
# change this your machine's ip, localhost won't work
# because localhost refers to the docker container.
- targets: ['172.20.239.48:2500']
- job_name: loki
static_configs:
- targets: ['loki:3100']
- job_name: prometheus
static_configs:
- targets: ['localhost:9090']
- job_name: minio
metrics_path: /minio/v2/metrics/cluster
static_configs:
- targets: ['minio:9000']

View File

@ -1,48 +0,0 @@
stream_over_http_enabled: true
server:
http_listen_port: 3200
log_level: info
query_frontend:
search:
duration_slo: 5s
throughput_bytes_slo: 1.073741824e+09
metadata_slo:
duration_slo: 5s
throughput_bytes_slo: 1.073741824e+09
trace_by_id:
duration_slo: 5s
distributor:
receivers:
otlp:
protocols:
grpc:
endpoint: "tempo:4317"
metrics_generator:
registry:
external_labels:
source: tempo
cluster: docker-compose
storage:
path: /var/tempo/generator/wal
remote_write:
- url: http://prometheus:9090/api/v1/write
send_exemplars: true
traces_storage:
path: /var/tempo/generator/traces
storage:
trace:
backend: local # backend configuration to use
wal:
path: /var/tempo/wal # where to store the wal locally
local:
path: /var/tempo/blocks
overrides:
defaults:
metrics_generator:
processors: [service-graphs, span-metrics, local-blocks] # enables metrics generator
generate_native_histograms: both

Binary file not shown.

Before

Width:  |  Height:  |  Size: 264 KiB

233
src/db.rs
View File

@ -1,39 +1,59 @@
use base64::{
alphabet,
engine::{self, general_purpose},
Engine,
};
use metrics::counter;
use serde::{Deserialize, Serialize};
use std::{fmt::Debug, time::Duration};
use serde::{ser::SerializeStruct, Deserialize, Serialize};
use std::{collections::HashSet, fmt::Debug, sync::LazyLock, time::Instant};
use surrealdb::{
engine::remote::ws::{Client, Ws},
opt::auth::Root,
sql::Thing,
Surreal,
Error::Api,
Response, Surreal,
};
use tokio::time::sleep;
use tracing::{error, instrument, trace};
use tokio::sync::Mutex;
use tracing::{error, instrument, trace, warn};
use url::Url;
use crate::Config;
use crate::{Config, Timer};
const STORE: &str = "surql_store_calls";
// static LOCK: LazyLock<Arc<Mutex<bool>>> = LazyLock::new(|| Arc::new(Mutex::new(true)));
static LOCK: LazyLock<Mutex<bool>> = LazyLock::new(|| Mutex::new(true));
#[derive(Serialize, Deserialize, Clone, Eq, PartialEq, Hash)]
const CUSTOM_ENGINE: engine::GeneralPurpose =
engine::GeneralPurpose::new(&alphabet::URL_SAFE, general_purpose::NO_PAD);
const TIME_SPENT_ON_LOCK: &'static str = "surql_lock_waiting_ms";
const STORE: &'static str = "surql_store_calls";
const LINK: &'static str = "surql_link_calls";
#[derive(Deserialize, Clone, Hash, Eq, PartialEq)]
pub struct Website {
pub id: Option<Thing>,
/// The url that this data is found at
pub site: Url,
/// Wether or not this link has been crawled yet
pub crawled: bool,
/// 200, 404, etc
pub status_code: u16,
}
impl Serialize for Website {
fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>
where
S: serde::Serializer,
{
let mut state = serializer.serialize_struct("Website", 2)?;
state.serialize_field("crawled", &self.crawled)?;
// to_string() calls the correct naming of site
state.serialize_field("site", &self.site.to_string())?;
state.end()
}
}
// manual impl to make tracing look nicer
impl Debug for Website {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
f.debug_struct("Website")
.field("host", &self.site.host())
.field("path", &self.site.path())
.field("status_code", &self.status_code)
.finish()
f.debug_struct("Website").field("site", &self.site).finish()
}
}
@ -44,88 +64,156 @@ impl Website {
Ok(a) => a,
Err(_) => todo!(),
};
Self {
crawled,
site,
status_code: 0,
id: None,
Self { crawled, site }
}
pub fn set_crawled(&mut self) {
trace!("Set crawled to true");
self.crawled = true
}
pub fn get_url_as_string(site: &Url) -> String {
let domain = match site.domain() {
Some(s) => s.to_string(),
None => {
warn!(
"Failed to get domain of URL: {}, falling back to 'localhost'",
site.to_string()
);
"localhost".to_string()
}
};
let path = site.path();
domain + path
}
pub fn get_url_as_b64_path(site: &Url) -> String {
let domain = site.domain().unwrap_or("DOMAIN").to_string();
let path = &CUSTOM_ENGINE.encode(site.path());
domain + path
}
#[instrument(skip_all)]
pub async fn links_to(&self, other: Vec<Thing>, db: &Surreal<Client>) {
let len = other.len();
if len == 0 {
return;
}
let from = &self.site;
// let to = other.site.to_string();
trace!("Linking {} pages to {from}", other.len());
let msg = format!("Linked {len} pages to {from}");
let timer = Timer::start(&msg);
// prevent the timer from being dropped instantly.
let _ = timer;
counter!(LINK).increment(1);
match db
.query("COUNT(RELATE (SELECT id FROM website WHERE site = $in) -> links_to -> $out)")
.bind(("in", from.clone()))
.bind(("out", other))
.await
{
Ok(mut e) => {
// The relate could technically "fail" (not relate anything), this just means that
// the query was ok.
let _: Response = e;
if let Ok(vec) = e.take(0) {
let _: Vec<usize> = vec;
if let Some(num) = vec.get(0) {
if *num == len {
trace!("Link for {from} OK - {num}/{len}");
return;
} else {
error!(
"Didn't link all the records. {num}/{len}. Surreal response: {:?}",
e
);
return;
}
}
}
error!("Linking request succeeded but couldn't verify the results.");
}
Err(e) => {
error!("{}", e.to_string());
}
}
}
pub async fn store_self(&self, db: &Surreal<Client>) {
counter!(STORE).increment(1);
db.query(
"INSERT INTO website $self
ON DUPLICATE KEY UPDATE
crawled = crawled OR $input.crawled
RETURN VALUE id;
",
)
.await
.expect("Failed to store self");
}
// Insert ever item in the vec into surreal, crawled state will be preserved as TRUE
// if already in the database as such or incoming data is TRUE.
#[instrument(skip(db))]
pub async fn store_all(all: Vec<Self>, db: &Surreal<Client>) -> Vec<Thing> {
pub async fn store_all(all: HashSet<Self>, db: &Surreal<Client>) -> Vec<Thing> {
// NOTES:
// * all incoming Websites come in as !crawled
// * there are potentially duplicates in all
counter!(STORE).increment(1);
let mut things = Vec::with_capacity(all.len());
let now = Instant::now();
let lock = LOCK.lock().await;
counter!(TIME_SPENT_ON_LOCK).increment(now.elapsed().as_millis() as u64);
let mut results = Vec::with_capacity(all.len());
match db
.query(
// TODO making this an upsert would make sense, but
// upserts seem to be broken.
//
// Doesn't look like upsert can take in an array, so insert
// it is...
//
"INSERT INTO website $array
ON DUPLICATE KEY UPDATE
accessed_at = time::now(),
status_code = $input.status_code,
processing = false,
crawled = crawled OR $input.crawled
RETURN VALUE id;
",
last_write = time::now()
RETURN VALUE id;"
,
)
.bind(("array", all))
.await
{
Ok(mut id) => match id.take::<Vec<Thing>>(0) {
Ok(mut x) => things.append(&mut x),
Err(err) => error!("{:?}", err),
},
Err(err) => {
error!("{:?}", err);
}
}
things
}
}
/// Returns uncrawled links
#[instrument(skip(db, config))]
pub async fn get_next(db: &Surreal<Client>, config: &Config) -> Option<Website> {
let mut res: Option<Website> = None;
let mut fails = 0;
while res == None {
let mut response = db
.query("fn::get_next($format)")
.bind(("format", config.crawl_filter.to_string()))
.await
.expect("Hard-coded query failed..?");
res = match response.take(0) {
Ok(ok) => ok,
Err(_err) => {
// basically just CSMA/CA
let delay = rand::random_range(10..500);
sleep(Duration::from_millis(delay)).await;
fails += 1;
// Don't get stuck here forever, failing...
// (most I've seen is 1)
if fails > 5 {
error!("Max attempts to get_next() reached... ({fails})");
return None
Ok(mut x) => results.append(&mut x),
Err(err) => {
error!("{:?}", err);
}
None
}
};
},
Err(err) => error!("{:?}", err),
}
drop(lock);
results
}
res
}
#[derive(Debug, Serialize)]
#[allow(dead_code)]
pub struct Email {
pub email: String,
pub on: String,
}
#[derive(Debug, Deserialize)]
pub struct Record {
#[allow(dead_code)]
pub id: Thing,
}
#[instrument(skip_all, name = "SurrealDB")]
pub async fn connect(config: &Config) -> surrealdb::Result<Surreal<Client>> {
trace!("Establishing connection to surreal...");
@ -154,3 +242,4 @@ pub async fn connect(config: &Config) -> surrealdb::Result<Surreal<Client>> {
Ok(db)
}

View File

@ -1,77 +0,0 @@
use std::{io::ErrorKind, path::PathBuf};
use reqwest::header::HeaderValue;
use tokio::fs;
use tracing::{error, trace, warn};
use url::Url;
pub fn as_path(url: &Url, content_type: &HeaderValue) -> PathBuf {
// extract data from url to save it accurately
let mut url_path = PathBuf::from("./downloaded/".to_string() + url.domain().unwrap_or("UnknownDomain") + url.path());
if let Ok(header) = content_type.to_str() {
// text/html; charset=UTF-8; option=value
let ttype = if let Some((t, _)) = header.split_once(';') {
t
} else {
header
};
if let Some((ttype, subtype)) = ttype.split_once('/') {
trace!(url = url.to_string(), main_type = ttype, sub_type = subtype, "Found Content-Type to be: {ttype}/{subtype}");
// If the Content-Type header is "*/html" (most likely "text/html") and the path's
// extension is anything but html:
if subtype=="html" && !url_path.extension().is_some_and(|f| f=="html" || f=="htm" ) {
// time to slap a index.html to the end of that path there!
url_path = url_path.join("index.html");
}
}
} else {
warn!("Header: {:?} couldn't be parsed into a string!", content_type);
}
trace!(url = url.to_string(), path = &*url_path.to_string_lossy(), "Converted URL into path");
url_path
}
pub async fn init(filename: &PathBuf) -> Option<fs::File> {
let file = async || tokio::fs::OpenOptions::new()
.write(true)
.append(true)
.create(true)
.open(&filename).await;
match file().await {
Ok(ok) => {
trace!("Initialized file {}", filename.to_str().unwrap_or("N/A"));
Some(ok)
},
Err(err) => {
// the file/folder isn't found
if err.kind() == ErrorKind::NotFound {
if let Some(parent ) = &filename.parent() {
// create the folders
if let Err(err) = fs::create_dir_all(&parent).await {
error!("Dir creation: {err} {:?}", filename);
} else if let Ok(ok) = file().await {
return Some(ok);
}
} else {
error!("Couldn't get file's parents: {:?}", &filename);
}
} else if err.kind() == ErrorKind::NotADirectory {
// Example:
// 1. example.com/user
// 2. example.com/user/post
// If file 1 exists it will prevent file 2 from existing
// FIXME
error!("One of the parent directories is actually a file...")
} else {
error!("File open error: {err} {:?}", filename);
}
// we don't care about other errors, we can't/shouldn't fix them
None
}
}
}

View File

@ -1,316 +1,58 @@
#![feature(ip_from)]
#![feature(path_add_extension)]
#![deny(clippy::unwrap_used)]
extern crate html5ever;
use std::{
collections::HashSet, fs::File, io::Read, sync::{Arc, LazyLock}
collections::HashSet, fs::File, io::Read, net::{IpAddr, Ipv4Addr}, time::Instant
};
use futures_util::StreamExt;
use opentelemetry::{global::{self}, metrics::{Counter, Meter, UpDownCounter}};
use opentelemetry_otlp::{Protocol, WithExportConfig};
use db::{connect, Website};
use opentelemetry_sdk::{metrics::SdkMeterProvider, trace::SdkTracerProvider};
use metrics::{counter, gauge};
use metrics_exporter_prometheus::PrometheusBuilder;
use s3::S3;
use serde::Deserialize;
use surrealdb::{engine::remote::ws::Client, Surreal};
use tokio::{io::{AsyncWriteExt, BufWriter}, sync::RwLock, task::JoinSet};
use tracing::{debug, error, info, instrument, level_filters::LevelFilter, trace, warn};
use tokio::task::JoinSet;
use tracing::{debug, error, info, instrument, trace, trace_span, warn};
use tracing_subscriber::{fmt, layer::SubscriberExt, EnvFilter, Layer, Registry};
use crate::db::get_next;
mod db;
mod filesystem;
mod parser;
mod s3;
static METER: LazyLock<Meter> = LazyLock::new(|| global::meter("Internet_Mapper"));
static BEING_PROCESSED: LazyLock<UpDownCounter<i64>> = LazyLock::new(||
METER
.i64_up_down_counter("crawler_pages_being_processed")
.build()
);
static BEING_PARSED: LazyLock<UpDownCounter<i64>> = LazyLock::new(||
METER
.i64_up_down_counter("crawler_pages_being_parsed")
.build()
);
static BEING_STREAMED: LazyLock<UpDownCounter<i64>> = LazyLock::new(||
METER
.i64_up_down_counter("crawler_pages_being_streamed")
.build()
);
static GET_IN_FLIGHT: LazyLock<UpDownCounter<i64>> = LazyLock::new(||
METER
.i64_up_down_counter("crawler_gets_in_flight")
.build()
);
static TOTAL_BYTES_DOWN: LazyLock<Counter<u64>> = LazyLock::new(||
METER
.u64_counter("crawler_total_bytes_down")
.build()
);
static SITES_CRAWLED: LazyLock<Counter<u64>> = LazyLock::new(||
METER
.u64_counter("crawler_total_sites_crawled")
.build()
);
static CONFIG: LazyLock<Config> = LazyLock::new(|| {
let mut file = File::open("./Crawler.toml").expect("Failed to read Crawler.toml");
let mut buf = String::new();
let _ = file.read_to_string(&mut buf);
let config: Config = toml::from_str(&buf).expect("Failed to parse Crawler.toml");
config
});
// FIXME Traces aren't working on multiple threads, they block
// static TRACER: LazyLock<BoxedTracer> = LazyLock::new(|| global::tracer("Internet_Mapper"));
const GET_METRIC: &'static str = "total_gets";
const GET_IN_FLIGHT: &'static str = "gets_in_flight";
const SITES_CRAWLED: &'static str = "pages_crawled";
const BEING_PROCESSED: &'static str = "pages_being_processed";
#[derive(Deserialize)]
struct Config {
tracing_endpoint: String,
metrics_endpoint: String,
log_file: String,
surreal_ns: String,
surreal_db: String,
surreal_url: String,
surreal_username: String,
surreal_password: String,
s3_url: String,
s3_bucket: String,
s3_access_key: String,
s3_secret_key: String,
crawl_filter: String,
start_url: String,
budget: usize,
batch_size: usize,
}
#[tokio::main]
async fn main() {
println!("Logs and metrics are provided to the Grafana dashboard");
let total_runtime = Timer::start("Completed");
// Start TRACE / LOGGING / METRICS
load_logging(&CONFIG); // this seems to be working ok
global::set_tracer_provider(load_tracing(&CONFIG));
global::set_meter_provider(load_metrics(&CONFIG));
// When getting uncrawled pages, name must contain this variable. "" will effectively get ignored.
// let crawl_filter = "en.wikipedia.org/";
// let budget = 50;
let crawled = Arc::new(RwLock::new(0));
let starting_url = &CONFIG.start_url;
let db = connect(&CONFIG)
.await
.expect("Failed to connect to surreal, aborting.");
let reqwest = reqwest::Client::builder()
// .use_rustls_tls()
.gzip(true)
.build()
.expect("Failed to build reqwest client.");
// Kick off the whole machine - This Website object doesn't matter, it's just to allow for
// get() to work.
// let mut span = TRACER.start("Pre-Loop");
let site = Website::new(starting_url, false);
process(site, db.clone(), reqwest.clone()).await;
// span.end();
// let mut main_loop_span= TRACER.start("Main-Loop");
let mut futures = JoinSet::new();
for _ in 0..CONFIG.batch_size {
futures.spawn(process_single_thread(&CONFIG, db.clone(), reqwest.clone(), crawled.clone()));
}
futures.join_all().await;
// main_loop_span.end();
info!("Done");
}
async fn process_single_thread(config: &Config, db: Surreal<Client>, reqwest: reqwest::Client, crawled: Arc<RwLock<usize>>) {
while *(crawled.read().await) < config.budget {
let uncrawled = get_next(&db.clone(), &config).await;
match uncrawled {
Some(site) => {
process(site, db.clone(), reqwest.clone()).await;
SITES_CRAWLED.add(1, &[]);
// Somehow this write doesn't hang on the while's read?
let mut c = crawled.write().await;
*c += 1;
},
None => {
warn!("fn::get_next() returned None");
return;
},
}
}
}
#[instrument(skip(db, reqwest))]
/// Downloads and crawls and stores a webpage.
/// It is acceptable to clone `db`, `reqwest`, and `s3` because they all use `Arc`s internally. - Noted by Oliver
async fn process(mut site: Website, db: Surreal<Client>, reqwest: reqwest::Client) {
// METRICS
debug!(url = &site.site.as_str(), "Process: {}", &site.site);
BEING_PROCESSED.add(1, &[]);
// let mut process_span = TRACER.start("Process");
// Build the request
let request_builder = reqwest.get(site.site.to_string());
// Send the http request (get)
GET_IN_FLIGHT.add(1, &[]);
if let Ok(response) = request_builder.send().await {
GET_IN_FLIGHT.add(-1, &[]);
let headers = response.headers();
let code = response.status();
if code != 200 {
warn!("{code} for {}", site.site.as_str());
}
#[allow(non_snake_case)]
let CT = headers.get("Content-Type");
let ct = headers.get("content-type");
let ct = match (CT,ct) {
(None, None) => {
warn!("Server did not respond with Content-Type header. Url: {} Headers: ({:?})", site.site.to_string(), headers);
return
},
(None, Some(a)) => a,
(Some(a), None) => a,
(Some(a), Some(_)) => a,
};
// create filepath (handles / -> /index.html)
let path = filesystem::as_path(&site.site, ct);
let mut tmp_path= path.clone();
if !(tmp_path.add_extension("crawl_temp")) {
warn!("Failed to add extension to file");
// fallback ig
tmp_path = tmp_path.with_extension("crawl_temp");
}
// make sure that the file is good to go
if let Some(file) = filesystem::init(&tmp_path).await {
// Get body from response
// stream the response onto the disk
let mut stream = response.bytes_stream();
let should_parse = path.to_string_lossy().ends_with(".html");
let mut writer = BufWriter::new(file);
let mut buf: Vec<u8> = Vec::new();
// Write file to disk
trace!("Writing at: {:?}", tmp_path);
BEING_STREAMED.add(1, &[]);
// let mut stream_span = TRACER.start("Stream");
while let Some(data) = stream.next().await {
match data {
Ok(data) => {
TOTAL_BYTES_DOWN.add(data.len() as u64, &[]);
let _ = writer.write_all(&data).await;
// If we are going to parse this file later, we will save it
// into memory as well as the disk.
// We do this because the data here might be incomplete
if should_parse {
data.iter().for_each(|f| buf.push(*f));
}
},
Err(err) => {
error!("{err}")
},
}
}
let _ = writer.flush().await;
// rename the temp file into the real file name
if let Err(err) = tokio::fs::rename(&tmp_path, &path).await {
error!(
from = &*tmp_path.to_string_lossy(),
to = &*path.to_string_lossy(),
"Error renaming file: {}",
err
);
}
// stream_span.end();
BEING_STREAMED.add(-1, &[]);
// (If needed) Parse the file
if should_parse {
BEING_PARSED.add(1, &[]);
// let mut parsing_span = TRACER.start("Parsing");
// Parse document and get relationships
let sites = parser::parse(&site, &buf).await;
// De-duplicate this list
let prev_len = sites.len();
let set = sites.into_iter().fold(HashSet::new(), |mut set, item| {
set.insert(item);
set
});
let de_dupe_sites: Vec<Website> = set.into_iter().collect();
let diff = prev_len - de_dupe_sites.len();
trace!("Saved {diff} from being entered into the db by de-duping");
// Store all the other sites so that we can link to them.
let _ = Website::store_all(de_dupe_sites, &db).await;
// parsing_span.end();
BEING_PARSED.add(-1, &[]);
} else {
trace!(url = site.site.as_str(), "Parse = False");
}
// update self in db
site.crawled = true;
site.status_code = code.as_u16();
Website::store_all(vec![site.clone()], &db).await;
}
} else {
error!(url = site.site.as_str(), "Failed to get: {}", &site.site);
}
// process_span.end();
BEING_PROCESSED.add(-1, &[]);
}
fn load_tracing(config: &Config) -> SdkTracerProvider {
// Send spans to Alloy (which will send them to Tempo)
let otlp_span = opentelemetry_otlp::SpanExporter::builder()
.with_tonic()
.with_endpoint(config.tracing_endpoint.clone())
.build()
.unwrap();
let tracer_provider = opentelemetry_sdk::trace::SdkTracerProvider::builder()
.with_simple_exporter(otlp_span)
.build();
tracer_provider
}
fn load_logging(config: &Config) {
// let otlp_log = opentelemetry_otlp::LogExporter::builder()
// .with_tonic()
// .with_endpoint(endpoint)
// .build()
// .unwrap();
// let logger_provider = opentelemetry_sdk::logs::SdkLoggerProvider::builder()
// .with_simple_exporter(otlp_log)
// .build();
let writer = std::fs::OpenOptions::new()
.append(true)
.create(true)
.open(config.log_file.clone())
.open("./docker/logs/tracing.log")
.expect("Couldn't make log file!");
let filter = EnvFilter::builder()
.with_default_directive(LevelFilter::DEBUG.into())
.from_env_lossy();
let filter = EnvFilter::from_default_env();
let registry = Registry::default().with(
fmt::layer()
@ -323,20 +65,206 @@ fn load_logging(config: &Config) {
);
tracing::subscriber::set_global_default(registry).expect("Failed to set default subscriber");
}
fn load_metrics(config: &Config) -> SdkMeterProvider {
// Send metrics to Prometheus
let otlp_metrics = opentelemetry_otlp::MetricExporter::builder()
.with_http()
.with_protocol(Protocol::HttpBinary)
.with_endpoint(config.metrics_endpoint.clone())
let builder = PrometheusBuilder::new();
builder
.with_http_listener(std::net::SocketAddr::new(
IpAddr::V4(Ipv4Addr::from_octets([0, 0, 0, 0])),
2500,
))
.install()
.expect("failed to install recorder/exporter");
debug!("Starting...");
// Would probably take these in as parameters from a cli
let starting_url = "https://en.wikipedia.org/";
// When getting uncrawled pages, name must contain this variable. "" will effectively get ignored.
// let crawl_filter = "en.wikipedia.org/";
// let budget = 50;
let mut crawled = 0;
let mut file = File::open("./Crawler.toml").expect("Failed to read Crawler.toml");
let mut buf = String::new();
let _ = file.read_to_string(&mut buf);
let config: Config = toml::from_str(&buf).expect("Failed to parse Crawler.toml");
let db = connect(&config)
.await
.expect("Failed to connect to surreal, aborting.");
let s3 = S3::connect(&config)
.await
.expect("Failed to connect to minio, aborting.\n\nThis probably means you need to login to the minio console and get a new access key!\n\n(Probably here) http://localhost:9001/access-keys/new-account\n\n");
let reqwest = reqwest::Client::builder()
// .use_rustls_tls()
.gzip(true)
.build()
.unwrap();
let metrics_provider = opentelemetry_sdk::metrics::SdkMeterProvider::builder()
.with_periodic_exporter(otlp_metrics) // default delay is 60s, turn down to like 15
.build();
metrics_provider
.expect("Failed to build reqwest client.");
// Kick off the whole machine - This Website object doesn't matter, it's just to allow for
// get() to work.
let span = trace_span!("Pre-Loop");
let pre_loop_span = span.enter();
// Download the site
let site = Website::new(&starting_url, false);
process(site, db.clone(), reqwest.clone(), s3.clone()).await;
drop(pre_loop_span);
let span = trace_span!("Loop");
let span = span.enter();
while crawled < config.budget {
let get_num = if config.budget - crawled < 100 {
config.budget - crawled
} else {
100
};
let uncrawled = get_uncrawled_links(&db, get_num, config.crawl_filter.clone()).await;
if uncrawled.len() == 0 {
info!("Had more budget but finished crawling everything.");
return;
}
debug!("Crawling {} pages...", uncrawled.len());
let span = trace_span!("Crawling");
let _ = span.enter();
{
let mut futures = JoinSet::new();
for site in uncrawled {
gauge!(BEING_PROCESSED).increment(1);
futures.spawn(process(site, db.clone(), reqwest.clone(), s3.clone()));
// let percent = format!("{:.2}%", (crawled as f32 / budget as f32) * 100f32);
// info!("Crawled {crawled} out of {budget} pages. ({percent})");
}
debug!("Joining {} futures...", futures.len());
let c = counter!(SITES_CRAWLED);
// As futures complete runs code in while block
while let Some(_) = futures.join_next().await {
c.increment(1);
gauge!(BEING_PROCESSED).decrement(1);
crawled += 1;
}
}
}
drop(span);
debug!("Done");
drop(total_runtime);
}
#[instrument(skip(db, s3, reqwest))]
/// Downloads and crawls and stores a webpage.
/// It is acceptable to clone `db`, `reqwest`, and `s3` because they all use `Arc`s internally. - Noted by Oliver
async fn process(mut site: Website, db: Surreal<Client>, reqwest: reqwest::Client, s3: S3) {
// METRICS
trace!("Process: {}", &site.site);
let timer = Timer::start("Built request");
// Build the request
let request_builder = reqwest.get(&site.site.to_string());
// METRICS
timer.stop();
// METRICS
let g = gauge!(GET_IN_FLIGHT);
g.increment(1);
counter!(GET_METRIC).increment(1);
let timer = Timer::start("Got page");
// Send the http request (get)
if let Ok(response) = request_builder.send().await {
// METRICS
timer.stop();
g.decrement(1);
// Get body from response
let data = response
.text()
.await
.expect("Failed to read http response's body!");
// Store document
s3.store(&data, &site.site).await;
// Parse document and get relationships
let sites = parser::parse(&site, &data).await;
// update self in db
site.set_crawled();
site.store_self(&db).await;
// de duplicate this list
let set = sites.iter().fold(HashSet::new(), |mut set, item| {
// TODO seems kinda dumb to clone everything.
set.insert(item.clone());
set
});
trace!("Shrunk items to store from {} to {}", sites.len(), set.len());
// Store all the other sites so that we can link to them.
let others = Website::store_all(set, &db).await;
// Make the database's links reflect the html links between sites
site.links_to(others, &db).await;
} else {
error!("Failed to get: {}", &site.site);
}
}
/// Returns uncrawled links
#[instrument(skip(db))]
async fn get_uncrawled_links(
db: &Surreal<Client>,
mut count: usize,
filter: String,
) -> Vec<Website> {
if count > 100 {
count = 100
}
debug!("Getting uncrawled links");
let mut response = db
.query("SELECT * FROM website WHERE crawled = false AND site ~ type::string($format) LIMIT $count;")
.bind(("format", filter))
.bind(("count", count))
.await
.expect("Hard-coded query failed..?");
response
.take(0)
.expect("Returned websites couldn't be parsed")
}
pub struct Timer<'a> {
start: Instant,
msg: &'a str,
}
impl<'a> Timer<'a> {
#[inline]
pub fn start(msg: &'a str) -> Self {
Self {
start: Instant::now(),
msg,
}
}
pub fn stop(&self) -> f64 {
let dif = self.start.elapsed().as_micros();
let ms = dif as f64 / 1000.;
if ms > 200. {
warn!("{}", format!("{} in {:.3}ms", self.msg, ms));
}
ms
}
}
impl Drop for Timer<'_> {
fn drop(&mut self) {
self.stop();
}
}

View File

@ -1,13 +1,14 @@
use std::default::Default;
use std::str::FromStr;
use html5ever::tokenizer::{BufferQueue, TokenizerResult};
use html5ever::tokenizer::{StartTag, TagToken};
use html5ever::tokenizer::{Token, TokenSink, TokenSinkResult, Tokenizer, TokenizerOpts};
use html5ever::{local_name, tendril::*};
use tracing::{error, instrument, trace, warn};
use url::Url;
use tracing::instrument;
use crate::db::Website;
use crate::Timer;
impl TokenSink for Website {
type Handle = Vec<Website>;
@ -17,7 +18,6 @@ impl TokenSink for Website {
TagToken(tag) => {
if tag.kind == StartTag {
match tag.name {
// this should be all the html elements that have links
local_name!("a")
| local_name!("audio")
| local_name!("area")
@ -32,18 +32,23 @@ impl TokenSink for Website {
let attr_name = attr.name.local.to_string();
if attr_name == "src" || attr_name == "href" || attr_name == "data"
{
trace!(url = self.site.as_str(),"Found `{}` in html `{}` tag", &attr.value, tag.name);
let url = try_get_url(&self.site, &attr.value);
// Get clone of the current site object
let mut web = self.clone();
if let Some(mut parsed) = url {
parsed.set_query(None);
parsed.set_fragment(None);
trace!(url = self.site.as_str(), "Final cleaned URL: `{}`", parsed.to_string());
let web = Website::new(&parsed.to_string(), false);
links.push(web);
}
// Set url
let mut url = web.site;
url.set_fragment(None); // removes #xyz
let joined = url
.join(&attr.value)
.expect("Failed to join url during parsing!");
web.site = joined;
web.crawled = false;
links.push(web);
}
}
return TokenSinkResult::Script(links);
}
local_name!("button") | local_name!("meta") | local_name!("iframe") => {
@ -59,92 +64,32 @@ impl TokenSink for Website {
}
}
#[instrument(skip(data))]
#[instrument(skip_all)]
/// Parses the passed site and returns all the sites it links to.
pub async fn parse(site: &Website, data: &[u8]) -> Vec<Website> {
trace!(url = site.site.as_str(), "Parsing {}", site.site.to_string());
pub async fn parse(site: &Website, data: &str) -> Vec<Website> {
// prep work
let mut other_sites: Vec<Website> = Vec::new();
let _t = Timer::start("Parsed page");
// change data into something that can be tokenized
let s: Result<Tendril<fmt::UTF8>, ()> = Tendril::try_from_byte_slice(data);
if let Ok(chunk) = s {
// create buffer of tokens and push our input into it
let token_buffer = BufferQueue::default();
token_buffer.push_back(
chunk
.try_reinterpret::<fmt::UTF8>()
.expect("Failed to reinterpret chunk!"),
);
// create the tokenizer
let tokenizer = Tokenizer::new(site.clone(), TokenizerOpts::default());
let chunk = Tendril::from_str(&data).expect("Failed to parse string into Tendril!");
// create buffer of tokens and push our input into it
let mut token_buffer = BufferQueue::default();
token_buffer.push_back(
chunk
.try_reinterpret::<fmt::UTF8>()
.expect("Failed to reinterprt chunk!"),
);
// create the tokenizer
let tokenizer = Tokenizer::new(site.clone(), TokenizerOpts::default());
// go thru buffer
while let TokenizerResult::Script(mut sites) = tokenizer.feed(&token_buffer) {
other_sites.append(&mut sites);
// other_sites.push(sites);
}
assert!(token_buffer.is_empty());
tokenizer.end();
} else {
warn!(url = site.site.as_str(), "Tendril failed to parse on: {}", site.site.to_string());
// go thru buffer
while let TokenizerResult::Script(mut sites) = tokenizer.feed(&mut token_buffer) {
other_sites.append(&mut sites);
// other_sites.push(sites);
}
assert!(token_buffer.is_empty());
tokenizer.end();
other_sites
}
#[instrument]
fn try_get_url(parent: &Url, link: &str) -> Option<Url> {
match Url::parse(link) {
Ok(ok) => Some(ok),
Err(e) => {
if link.starts_with('#') {
trace!(url = parent.as_str(), "Rejecting # url");
None
} else if link.starts_with("//") {
// if a url starts with "//" is assumed that it will adopt
// the same scheme as it's parent
// https://stackoverflow.com/questions/9646407/two-forward-slashes-in-a-url-src-href-attribute
let scheme = parent.scheme();
match Url::parse(&format!("{scheme}://{link}")) {
Ok(url) => Some(url),
Err(err) => {
error!("Failed parsing relative scheme url: {}", err);
None
}
}
} else {
// # This is some sort of relative url, gonna try patching it up into an absolute
// url
match e {
url::ParseError::RelativeUrlWithoutBase => {
// Is: scheme://host:port
let mut origin = parent.origin().ascii_serialization();
if !origin.ends_with('/') && !link.starts_with('/') {
origin += "/";
}
let url = origin.clone() + link;
if let Ok(url) = Url::parse(&url) {
trace!(url = parent.as_str(), "Built `{url}` from `{origin} + `{}`", link.to_string());
Some(url)
} else {
error!(
"Failed to reconstruct a url from relative url: `{}` on site: `{}`. Failed url was: {}",
link,
parent.to_string(),
url
);
None
}
}
_ => {
error!("MISC error: {:?} {:?}", e, link);
None
}
}
}
}
}
}

101
src/s3.rs Normal file
View File

@ -0,0 +1,101 @@
use metrics::counter;
use minio::s3::{
args::{BucketExistsArgs, MakeBucketArgs},
client::ClientBuilder,
creds::StaticProvider,
error::Error,
http::BaseUrl,
Client,
};
use tracing::{instrument, trace, warn};
use url::Url;
use crate::{db::Website, Config, Timer};
const S3_ROUND_TRIP_METRIC: &'static str = "s3_trips";
#[derive(Clone)]
pub struct S3 {
bucket_name: String,
client: Client,
}
impl S3 {
#[instrument(skip_all, name = "S3")]
pub async fn connect(config: &Config) -> Result<Self, Error> {
let base_url = config
.s3_url
.parse::<BaseUrl>()
.expect("Failed to parse url into BaseUrl");
let static_provider =
StaticProvider::new(&config.s3_access_key, &config.s3_secret_key, None);
let client = ClientBuilder::new(base_url)
.provider(Some(Box::new(static_provider)))
.build()?;
trace!("Checking bucket...");
let exists = client
.bucket_exists(
&BucketExistsArgs::new(&config.s3_bucket)
.expect("Failed to check if bucket exists"),
)
.await?;
counter!(S3_ROUND_TRIP_METRIC).increment(1);
if !exists {
trace!("Creating bucket...");
client
.make_bucket(
&MakeBucketArgs::new(&config.s3_bucket).expect("Failed to create bucket!"),
)
.await?;
}
counter!(S3_ROUND_TRIP_METRIC).increment(1);
trace!("Connection successful");
Ok(Self {
bucket_name: config.s3_bucket.to_owned(),
client,
})
}
#[instrument(name = "s3_store", skip_all)]
pub async fn store(&self, data: &str, url: &Url) {
let counter = counter!(S3_ROUND_TRIP_METRIC);
let t = Timer::start("Stored page");
let _ = t; // prevent compiler drop
let filename = Website::get_url_as_string(url);
trace!("Storing {} as {filename}", url.to_string());
counter.increment(1);
let _ = match &self
.client
.put_object_content(&self.bucket_name, &filename, data.to_owned())
.send()
.await
{
Ok(_) => {}
Err(err) => match err {
Error::InvalidObjectName(_) => {
// This code will really only run if the url has non-english chars
warn!("Tried storing invalid object name, retrying with Base64 encoding. Last try.");
let filename: String = Website::get_url_as_b64_path(url);
counter.increment(1);
let _ = &self
.client
.put_object_content(&self.bucket_name, &filename, data.to_owned())
.send()
.await
.unwrap();
}
_ => {}
},
};
}
}

View File

@ -4,13 +4,5 @@ DEFINE FIELD IF NOT EXISTS site ON TABLE website TYPE string;
DEFINE INDEX IF NOT EXISTS idx ON TABLE website COLUMNS site UNIQUE;
DEFINE FIELD IF NOT EXISTS crawled ON TABLE website TYPE bool;
DEFINE FIELD IF NOT EXISTS processing ON TABLE website TYPE bool DEFAULT false;
DEFINE FIELD IF NOT EXISTS accessed_at ON TABLE website VALUE time::now();
DEFINE FIELD IF NOT EXISTS first_accessed_at ON TABLE website VALUE time::now();
DEFINE FUNCTION OVERWRITE fn::get_next($filter: string) {
LET $site = SELECT * FROM ONLY website WHERE crawled = false AND processing = false AND site ~ type::string($filter) LIMIT 1;
UPDATE $site.id SET processing = true;
RETURN $site
};
DEFINE FIELD IF NOT EXISTS created ON TABLE website VALUE time::now();