core: add support for sign text transformations

Add support for regexp replacement patterns, which can be useful when
the text matched by --sign-filter or --sign-prefix should not be
displayed.

src/core/common.rs

@@ -8,7 +8,7 @@ use std::{
 
 use anyhow::{Context, Result};
 use indexmap::IndexSet;
-use regex::RegexSet;
+use regex::{Regex, RegexSet};
 use serde::{Deserialize, Serialize};
 
 use crate::{
@@ -130,6 +130,7 @@ pub enum TileKind {
 }
 
 /// Common configuration based on command line arguments
+#[derive(Debug)]
 pub struct Config {
 	/// Number of threads for parallel processing
 	pub num_threads: usize,
@@ -151,6 +152,8 @@ pub struct Config {
 	pub viewer_entities_path: PathBuf,
 	/// Sign text filter patterns
 	pub sign_patterns: RegexSet,
+	/// Sign text transformation pattern
+	pub sign_transforms: Vec<(Regex, String)>,
 }
 
 impl Config {
@@ -173,6 +176,8 @@ impl Config {
 			.collect();
 
 		let sign_patterns = Self::sign_patterns(args).context("Failed to parse sign patterns")?;
+		let sign_transforms =
+			Self::sign_transforms(args).context("Failed to parse sign transforms")?;
 
 		Ok(Config {
 			num_threads,
@@ -185,6 +190,7 @@ impl Config {
 			viewer_info_path,
 			viewer_entities_path,
 			sign_patterns,
+			sign_transforms,
 		})
 	}
 
@@ -200,6 +206,28 @@ impl Config {
 		)?)
 	}
 
+	/// Parses the sign transform argument into a vector of [Regex] and
+	/// corresponding replacement strings
+	fn sign_transforms(args: &super::Args) -> Result<Vec<(Regex, String)>> {
+		let splitter = Regex::new(r"^s/((?:[^\\/]|\\.)*)/((?:[^\\/]|\\.)*)/$").unwrap();
+
+		args.sign_transform
+			.iter()
+			.map(|t| Self::sign_transform(&splitter, t))
+			.collect()
+	}
+
+	/// Parses the sign transform argument into a [Regex] and its corresponding
+	/// replacement string
+	fn sign_transform(splitter: &Regex, transform: &str) -> Result<(Regex, String)> {
+		let captures = splitter
+			.captures(transform)
+			.with_context(|| format!("Invalid transform pattern '{}'", transform))?;
+		let regexp = Regex::new(&captures[1])?;
+		let replacement = captures[2].to_string();
+		Ok((regexp, replacement))
+	}
+
 	/// Constructs the path to an input region file
 	pub fn region_path(&self, coords: TileCoords) -> PathBuf {
 		let filename = coord_filename(coords, "mca");

src/core/metadata_writer.rs

@@ -1,6 +1,7 @@
 //! The [MetadataWriter] and related types
 
 use anyhow::{Context, Result};
+use regex::Regex;
 use serde::Serialize;
 
 use crate::{
@@ -8,7 +9,7 @@ use crate::{
 	io::{fs, storage},
 	world::{
 		block_entity::{self, BlockEntity, BlockEntityData},
-		de,
+		de, sign,
 	},
 };
 
@@ -145,6 +146,28 @@ impl<'a> MetadataWriter<'a> {
 		false
 	}
 
+	/// Applies a single transform to a [sign::SignText]
+	///
+	/// The regular expression is applied for each line of the sign text
+	/// separately (actually for each element when JSON text is used)
+	fn sign_text_transform(sign_text: &mut sign::SignText, transform: &(Regex, String)) {
+		let (regexp, replacement) = transform;
+
+		for line in &mut sign_text.0 {
+			for text in &mut line.0 {
+				text.text = regexp.replace_all(&text.text, replacement).into_owned()
+			}
+		}
+	}
+
+	/// Applies the configured transforms to the text of a sign
+	fn sign_transform(&self, sign: &mut block_entity::Sign) {
+		for transform in &self.config.sign_transforms {
+			Self::sign_text_transform(&mut sign.front_text, transform);
+			Self::sign_text_transform(&mut sign.back_text, transform);
+		}
+	}
+
 	/// Generates [Entities] data from collected entity lists
 	fn entities(&self) -> Result<Entities> {
 		let data: ProcessedEntities =
@@ -158,6 +181,12 @@ impl<'a> MetadataWriter<'a> {
 				.filter(|entity| match &entity.data {
 					BlockEntityData::Sign(sign) => self.sign_filter(sign),
 				})
+				.map(|mut entity| {
+					match &mut entity.data {
+						BlockEntityData::Sign(sign) => self.sign_transform(sign),
+					};
+					entity
+				})
 				.collect(),
 		};
 

src/core/mod.rs

@@ -59,6 +59,12 @@ pub struct Args {
 	/// To make all signs visible, pass an empty string to either option.
 	#[arg(long)]
 	pub sign_filter: Vec<String>,
+	/// Regular expression replacement pattern for sign texts
+	///
+	/// Accepts patterns of the form 's/regexp/replacement/'. Transforms
+	/// are applied to each line of sign texts separately.
+	#[arg(long)]
+	pub sign_transform: Vec<String>,
 	/// Minecraft save directory
 	pub input_dir: PathBuf,
 	/// MinedMap data directory