@zzl221000 Thanks for the bug report.
Please help me test #158, if possible.

cargo run --bin caracal -- https://testfileorg.netwet.net/testfile.org-1GB-%20Corrupt.zip

@xrelkd Not downloading correctly

root@JimZhang:~/caracal# cargo run --bin caracal -- https://testfileorg.netwet.net/testfile.org-1GB-%20Corrupt.zip
    Finished dev [unoptimized + debuginfo] target(s) in 0.13s
     Running `target/debug/caracal 'https://testfileorg.netwet.net/testfile.org-1GB-%20Corrupt.zip'`
⠈ [00:00:35] [3/6 testfile.org-1GB-%20Corrupt.zip] [###########################] 37.15 MiB/s 100% 1.48 GiB/1.00 GiB (0s)

root@JimZhang:~/caracal# git branch -a
  develop
* fix/issue-157
  remotes/origin/HEAD -> origin/develop
  remotes/origin/develop
  remotes/origin/feat/yew
  remotes/origin/fix/issue-157
  remotes/origin/main

Alternatively, we can use hash to verify the correctness of the file.
After downloading with caracal, the hash of the file was changed.
I tried aria2 and it downloads the file correctly.

It seems that the https://testfileorg.netwet.net/ does not allow us to fetch partial content from it every time.

The server returns status code 200 instead of status code 206, when we try to fetch partial content from it with Range header sometimes.

[crates/engine/src/fetcher/http.rs:73:9] (start, end, content_length, resp.status(), resp.headers()) = (
    1073741820,
    1073741823,
    4,
    200,
    {
        "date": "Fri, 08 Mar 2024 15:54:45 GMT",
        "content-type": "application/x-zip-compressed",
        "content-length": "1073741824",
        "connection": "keep-alive",
        "etag": "\"e7cb7d707e841c450ce7396c75224528\"",
        "last-modified": "Fri, 06 Jan 2023 10:15:32 GMT",
        "vary": "Accept-Encoding",
        "cache-control": "max-age=14400",
        "cf-cache-status": "MISS",
        "report-to": "{\"endpoints\":[{\"url\":\"https:\/\/a.nel.cloudflare.com\/report\/v3?s=MvkvFsocSGqGSAk6iKw%2Bd6RFviLwbIaI5VaObCKfPsfCIFYGdJo%2Fqajgn37ysg31PYIW3ta6hJRppPecOPOxvWwASZXWVsNNn0vw573Q28pvEQkDuG4t0M0E0R9miNOkFN2bXAWTNsFO\"}],\"group\":\"cf-nel\",\"max_age\":604800}",
        "nel": "{\"success_fraction\":0,\"report_to\":\"cf-nel\",\"max_age\":604800}",
        "server": "cloudflare",
        "cf-ray": "8614016f9e5184db-HKG",
        "alt-svc": "h3=\":443\"; ma=86400",
    },
)
thread 'tokio-runtime-worker' panicked at crates/engine/src/fetcher/http.rs:74:9:
assertion `left == right` failed
  left: 200
 right: 206

You can use a single worker to download the file.
If you use cargo run --bin caracal -- -n 1 https://testfileorg.netwet.net/testfile.org-1GB-%20Corrupt.zip to download the file, the MD5 digest of the file is e7cb7d707e841c450ce7396c75224528.

We tested the server with wget, aria2, axel and caracal.
We found that wget and aria2 downloaded the file correctly, the MD5 digest is e7cb7d707e841c450ce7396c75224528.
Both wget and aria2 are using a single connection to fetch the file, even if you specify aria2 to use more than 2 connections.
axel and caracal downloaded the file with 6 connections, but the MD5 digest is not correct.

The only thing we can do is to refuse using multiple connections when the server does not guarantee use to fetch partial content every time.

Investigated aria2's issue and it runs into the same problem, just triggered differently.
aria2/aria2#1344 (comment)
Handling of RANGE responses more cumbersome than expected