feat(forge): Add regex_replace filter to support replacing text usi…

…ng regex. (#380)

CHANGELOG.md

@@ -8,6 +8,7 @@ What's changed
 
 * Add `enforce_trailing_dots` into the `comment_formats` configuration. ([#XXX](...) by lquerel).
 * Add support for `indent_type` in both the comment filter and the `comment_formats` configuration. ([#XXX](...) by lquerel).
+* Add `regex_replace` filter to support replacing text using regex. ([#XXX](...) by lquerel).
 
 ## [0.9.2] - 2024-09-09
 

crates/weaver_forge/README.md

@@ -486,6 +486,9 @@ The following filters are available:
 - `acronym`: Replaces acronyms in the input string with the full name defined in the `acronyms` section of the
   `weaver.yaml` configuration file.
 - `split_id`: Splits a string by '.' creating a list of nested ids.
+- `regex_replace`: Replace all occurrences of a regex pattern (1st parameter) in the input string with the replacement
+  string (2nd parameter). Under the hood, this filter uses the `regex` crate (see
+  [regex](https://docs.rs/regex/latest/regex/index.html#traits) for more details) 
 - `comment_with_prefix(prefix)`: Outputs a multiline comment with the given prefix. This filter is deprecated, please use the more general `comment` filter.
 - `comment`: A generic comment formatter that uses the `comment_formats` section of the `weaver.yaml` configuration file (more details [here](#comment-filter)).
 - `flatten`: Converts a List of Lists into a single list with all elements.  

crates/weaver_forge/src/extensions/util.rs

@@ -6,6 +6,7 @@ use crate::config::WeaverConfig;
 use minijinja::value::Rest;
 use minijinja::{Environment, ErrorKind, Value};
 use regex::Regex;
+use std::borrow::Cow;
 use std::collections::HashMap;
 use std::sync::OnceLock;
 
@@ -17,6 +18,7 @@ pub(crate) fn add_filters(env: &mut Environment<'_>, target_config: &WeaverConfi
     );
     env.add_filter("flatten", flatten);
     env.add_filter("split_id", split_id);
+    env.add_filter("regex_replace", regex_replace);
 }
 
 /// Add utility functions to the environment.
@@ -66,6 +68,24 @@ fn split_id(value: Value) -> Result<Vec<Value>, minijinja::Error> {
     }
 }
 
+/// Replace all occurrences of a regex pattern (1st parameter) in the input string with the
+/// replacement string (2nd parameter).
+fn regex_replace(
+    input: Cow<'_, str>,
+    pattern: Cow<'_, str>,
+    replacement: Cow<'_, str>,
+) -> Result<String, minijinja::Error> {
+    let re = Regex::new(pattern.as_ref()).map_err(|e| {
+        minijinja::Error::new(
+            ErrorKind::InvalidOperation,
+            format!("Invalid regex pattern: {}", e),
+        )
+    })?;
+    Ok(re
+        .replace_all(input.as_ref(), replacement.as_ref())
+        .to_string())
+}
+
 /// Create a filter that replaces acronyms in the input string with the full
 /// name defined in the `acronyms` list.
 ///
@@ -110,3 +130,33 @@ pub fn acronym(acronyms: Vec<String>) -> impl Fn(&str) -> String {
             .collect()
     }
 }
+
+#[cfg(test)]
+mod tests {
+    use crate::extensions::util::add_filters;
+    use minijinja::Environment;
+
+    #[test]
+    fn test_regex_replace() {
+        let mut env = Environment::new();
+        let ctx = serde_json::Value::Null;
+        let config = crate::config::WeaverConfig::default();
+
+        add_filters(&mut env, &config);
+
+        assert_eq!(
+            env.render_str("{{ 'Hello World!' | regex_replace('!','?') }}", &ctx)
+                .unwrap(),
+            "Hello World?"
+        );
+
+        assert_eq!(
+            env.render_str(
+                "{{ \"This a test with multiple a's\" | regex_replace('a','A') }}",
+                &ctx
+            )
+            .unwrap(),
+            "This A test with multiple A's"
+        );
+    }
+}