pemfile: also trim leading contiguous content whitespace

RFC 7468 §2[0] says:
> parsers SHOULD ignore whitespace

Previously we only stripped the trailing whitespace from the base64
content lines within the PEM section boundaries. This branch updates our
processing to additionally trim leading contiguous whitespace. This felt
more sensible than outright ignoring whitespace (e.g. from the middle of
content) and will be sufficient to resolve the real world PEM files
we've seen with leading whitespace.

We base our implementation on the stdlib's unstable `[u8]::trim_ascii`
fn, which we can't (yet) use directly due to our MSRV/stable rust
requirements. A TODO is left for future cleanup.

[0]: https://www.rfc-editor.org/rfc/rfc7468#section-2

src/pemfile.rs

@@ -215,14 +215,8 @@ fn read_one_impl(
     }
 
     if section.is_some() {
-        let mut trim = 0;
-        for &b in line.iter().rev() {
-            match b {
-                b'\n' | b'\r' | b' ' => trim += 1,
-                _ => break,
-            }
-        }
-        b64buf.extend(&line[..line.len() - trim]);
+        // Extend b64buf without leading or trailing whitespace
+        b64buf.extend(trim_ascii(line));
     }
 
     Ok(ControlFlow::Continue(()))
@@ -266,6 +260,35 @@ fn read_until_newline<R: io::BufRead + ?Sized>(
     }
 }
 
+/// Trim contiguous leading and trailing whitespace from `line`.
+///
+/// We use [u8::is_ascii_whitespace] to determine what is whitespace.
+// TODO(XXX): Replace with `[u8]::trim_ascii` once stabilized[0] and available in our MSRV.
+//   [0]: https://github.com/rust-lang/rust/issues/94035
+const fn trim_ascii(line: &[u8]) -> &[u8] {
+    let mut bytes = line;
+
+    // Note: A pattern matching based approach (instead of indexing) allows
+    // making the function const.
+    while let [first, rest @ ..] = bytes {
+        if first.is_ascii_whitespace() {
+            bytes = rest;
+        } else {
+            break;
+        }
+    }
+
+    while let [rest @ .., last] = bytes {
+        if last.is_ascii_whitespace() {
+            bytes = rest;
+        } else {
+            break;
+        }
+    }
+
+    bytes
+}
+
 /// Extract and return all PEM sections by reading `rd`.
 #[cfg(feature = "std")]
 pub fn read_all(rd: &mut dyn io::BufRead) -> impl Iterator<Item = Result<Item, io::Error>> + '_ {
@@ -284,3 +307,31 @@ mod base64 {
     );
 }
 use self::base64::Engine;
+
+#[cfg(test)]
+mod tests {
+    #[test]
+    fn test_trim_ascii() {
+        let tests: &[(&[u8], &[u8])] = &[
+            (b"", b""),
+            (b"   hello world   ", b"hello world"),
+            (b"   hello\t\r\nworld   ", b"hello\t\r\nworld"),
+            (b"\n\r  \ttest\t  \r\n", b"test"),
+            (b"   \r\n  ", b""),
+            (b"no trimming needed", b"no trimming needed"),
+            (
+                b"\n\n content\n\n more content\n\n",
+                b"content\n\n more content",
+            ),
+        ];
+
+        for &(input, expected) in tests {
+            assert_eq!(
+                super::trim_ascii(input),
+                expected,
+                "Failed for input: {:?}",
+                input,
+            );
+        }
+    }
+}

tests/data/whitespace-prefix.crt

@@ -0,0 +1,21 @@
+-----BEGIN CERTIFICATE-----
+ MIIDaTCCAlGgAwIBAgIJAOq/zL+84IswMA0GCSqGSIb3DQEBCwUAMFoxCzAJBgNV
+ BAYTAlVTMQswCQYDVQQIDAJOQzEMMAoGA1UEBwwDUlRQMQ8wDQYDVQQKDAZOZXRB
+ cHAxDTALBgNVBAsMBEVTSVMxEDAOBgNVBAMMB1NTRk1DQ0EwHhcNMTcxMTAxMjEw
+ OTQyWhcNMjcxMDMwMjEwOTQyWjBaMQswCQYDVQQGEwJVUzELMAkGA1UECAwCTkMx
+ DDAKBgNVBAcMA1JUUDEPMA0GA1UECgwGTmV0QXBwMQ0wCwYDVQQLDARFU0lTMRAw
+ DgYDVQQDDAdTU0ZNQ0NBMIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEA
+ iaD9Ee0Yrdka0I+9GTJBIW/Fp5JU6kyjaxfOldW/R9lEubegXQFhDD2Xi1HZ+fTM
+ f224glB9xLJXAHhipRK01C2MgC4kSH75WL1iAiYeOBloExqmK6OCX+sdyO7RXm/H
+ Ra9tN2INWdvyO2pnmxsSnq56mCMsUZLtrRKp89FWgcxLg5r8QxH7xwfh5k54rxjE
+ 144TD9yrIiQOgRSIRHUrVJ9l/F/gnwzP8wcNABeXwN71Mzl7mliPA703kONQIAyU
+ 0E0tLpmy/U8dZdMmTBZGB7jI9f95Hl1RunfwhR371a6z38kgkvwrLzl4qflfsPjw
+ K9n4omNk9rCH9H9tWkxxjwIDAQABozIwMDAdBgNVHQ4EFgQU/bFyCCnqdDFKlQBJ
+ ExtV6wcMYkEwDwYDVR0TAQH/BAUwAwEB/zANBgkqhkiG9w0BAQsFAAOCAQEAOQMs
+ Pz2iBD1+3RcSOsahB36WAwPCjgPiXXXpU+Zri11+m6I0Lq+OWtf+YgaQ8ylLmCQd
+ 0p1wHlYA4qo896SycrhTQfy9GlS/aQqN192k3oBGoJcMIUnGUBGuEvyZ2aDUfkzy
+ JUqBe+0KaT7pkvvbRL7VUz34I7ouq9fQIRZ26vUDLTY3KM1n/DXBj3e30GHGMV3K
+ NN2twuLXPNjnryfgpliHU1rwV7r1WvrCVn4StjimP2bO5HGqD/SbiYUL2M9LOuLK
+ 6mqY4OHumYXq3k7CHrvt0FepsN0L14LYEt1LvpPDFWP3SdN4z4KqT9AGqBaJnhhl
+ Qiq8GWnAChspdBLxCg==
+-----END CERTIFICATE-----

tests/integration.rs

@@ -196,3 +196,15 @@ fn different_line_endings() {
         assert!(matches!(cert, rustls_pemfile::Item::X509Certificate(_)));
     }
 }
+
+#[test]
+fn whitespace_prefix() {
+    let items = rustls_pemfile::read_all(&mut BufReader::new(
+        &include_bytes!("data/whitespace-prefix.crt")[..],
+    ))
+    .collect::<Result<Vec<_>, _>>()
+    .unwrap();
+
+    assert_eq!(items.len(), 1);
+    assert!(matches!(items[0], rustls_pemfile::Item::X509Certificate(_)));
+}