Add all commonmark tests

1 files changed, 88 insertions, 0 deletions

diff --git a/build.rs b/build.rs
index cdfeac4..feb22c8 100644
--- a/build.rs
+++ b/build.rs
@@ -1,8 +1,96 @@
 extern crate reqwest;
+use regex::Regex;
 use std::fs;
 
 #[tokio::main]
 async fn main() {
+    commonmark().await;
+    punctuation().await;
+}
+
+async fn commonmark() {
+    let url = "https://raw.githubusercontent.com/commonmark/commonmark-spec/0.30/spec.txt";
+    let data_url = "commonmark-data.txt";
+    #[allow(unused_variables)]
+    let code_url = "tests/commonmark.rs";
+
+    let value = if let Ok(value) = fs::read_to_string(data_url) {
+        value
+    } else {
+        let value = reqwest::get(url).await.unwrap().text().await.unwrap();
+
+        fs::write(data_url, value.clone()).unwrap();
+
+        value
+    };
+
+    let re = Regex::new(r"(?m)(?:^`{32} example\n[\s\S]*?\n`{32}$|^#{1,6} *(.*)$)").unwrap();
+    let re_heading_prefix = Regex::new(r"#{1,6} ").unwrap();
+    let re_in_out = Regex::new(r"\n\.(?:\n|$)").unwrap();
+    let mut current_heading: Option<String> = None;
+    let mut case_index = 0;
+
+    let value = Regex::new(r"<!-- END TESTS -->[\s\S]*")
+        .unwrap()
+        .replace(&value, "");
+    let value = Regex::new(r"→").unwrap().replace_all(&value, "\t");
+    let mut cases: Vec<String> = vec![];
+
+    for mat in re.find_iter(&value) {
+        let mut lines = mat.as_str().lines().collect::<Vec<_>>();
+
+        if lines.len() == 1 {
+            current_heading = Some(re_heading_prefix.replace(lines[0], "").to_string());
+        } else {
+            lines.remove(0);
+            lines.pop();
+            let section = current_heading.as_ref().unwrap();
+            let case = lines.join("\n");
+            let parts = re_in_out.split(&case).collect::<Vec<_>>();
+            let input = format!("{}\n", parts[0]);
+            let output = if parts[1].is_empty() {
+                "".to_string()
+            } else {
+                format!("{}\n", parts[1])
+            };
+
+            let test = format!("    assert_eq!(\n        micromark_with_options(r###\"{}\"###, DANGER),\n        r###\"{}\"###,\n        r###\"{} ({})\"###\n);", input, output, section, case_index);
+
+            cases.push(test);
+
+            case_index += 1;
+        }
+    }
+
+    #[allow(unused_variables)]
+    let doc = format!(
+        "//! CommonMark test suite.
+
+// > 👉 **Important**: this module is generated by `build.rs`.
+// > It is generate from the latest Unicode data.
+
+extern crate micromark;
+use micromark::{{micromark_with_options, Options}};
+
+const DANGER: &Options = &Options {{
+    allow_dangerous_html: true,
+    allow_dangerous_protocol: true,
+    default_line_ending: None,
+}};
+
+#[test]
+fn commonmark() {{
+{}
+}}
+",
+        cases.join("\n\n")
+    );
+
+    // To do: enable when CM is completely fixed.
+    // fs::write(code_url, doc).unwrap();
+}
+
+async fn punctuation() {
     let url = "https://www.unicode.org/Public/UCD/latest/ucd/UnicodeData.txt";
     let data_url = "unicode-data.txt";
     let code_url = "src/unicode.rs";