From 309b7583bd84ce2b45109a61b96139054bf8a577 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Fe=CC=81lix=20Saparelli?= <felix@passcod.name>
Date: Sat, 8 Apr 2023 20:20:20 +1200
Subject: [PATCH] Add read filter

---
 crates/cli/src/args.rs             |  7 ++--
 crates/cli/src/filterer/proglib.rs | 58 +++++++++++++++++++++++++++++-
 2 files changed, 62 insertions(+), 3 deletions(-)

diff --git a/crates/cli/src/args.rs b/crates/cli/src/args.rs
index bf15052c..88c18136 100644
--- a/crates/cli/src/args.rs
+++ b/crates/cli/src/args.rs
@@ -802,8 +802,11 @@ pub struct Args {
 	///
 	///   - 'path | metadata' returns file metadata or null if the file does not exist.
 	///
-	///   - 'path | read' and 'path read(bytes)' return a string containing the file at path, or
-	///     the first n bytes of it. Obviously care should be taken not to read large files.
+	///   - 'path | filesize' returns the size of the file at path, or null if it does not exist.
+	///
+	///   - 'path | read(bytes)' returns a string containing the first n bytes of the file at path.
+	///     If the file is smaller than n bytes, the whole file is returned. There is no filter to
+	///     read the whole file at once to encourage limiting the amount of data read and processed.
 	///
 	///   - 'string | hash', and 'path | hashfile' return the hash of the string or file at path.
 	///     No guarantee is made about the algorithm used: treat it as an opaque value.
diff --git a/crates/cli/src/filterer/proglib.rs b/crates/cli/src/filterer/proglib.rs
index 5a287753..5c721588 100644
--- a/crates/cli/src/filterer/proglib.rs
+++ b/crates/cli/src/filterer/proglib.rs
@@ -1,4 +1,9 @@
-use std::{iter::once, sync::Arc};
+use std::{
+	fs::File,
+	io::{BufReader, Read},
+	iter::once,
+	sync::Arc,
+};
 
 use dashmap::DashMap;
 use jaq_core::{CustomFilter, Definitions, Error, Val};
@@ -43,6 +48,17 @@ macro_rules! string_arg {
 	};
 }
 
+macro_rules! int_arg {
+	($args:expr, $n:expr, $ctx:expr, $val:expr) => {
+		match $args[$n].run(($ctx.clone(), $val.clone())).next() {
+			Some(Ok(Val::Int(v))) => Ok(v as _),
+			Some(Ok(val)) => custom_err(format!("expected int but got {val:?}")),
+			Some(Err(e)) => Err(e),
+			None => custom_err("value expected but none found"),
+		}
+	};
+}
+
 macro_rules! log_action {
 	($level:expr, $val:expr) => {
 		match $level.to_ascii_lowercase().as_str() {
@@ -170,5 +186,45 @@ pub fn load_watchexec_defs(defs: &mut Definitions) -> miette::Result<()> {
 		}),
 	);
 
+	trace!("jaq: add read filter");
+	defs.insert_custom(
+		"read",
+		CustomFilter::new(1, {
+			move |args, (ctx, val)| {
+				let path = match &val {
+					Val::Str(v) => v.to_string(),
+					_ => return_err!(custom_err("expected string (path) but got {val:?}")),
+				};
+
+				let bytes = match int_arg!(args, 0, ctx, &val) {
+					Ok(v) => v,
+					Err(e) => return_err!(Err(e)),
+				};
+
+				Box::new(once(Ok(match File::open(&path) {
+					Ok(file) => {
+						let buf_reader = BufReader::new(file);
+						let mut limited = buf_reader.take(bytes);
+						let mut buffer = String::with_capacity(bytes as _);
+						match limited.read_to_string(&mut buffer) {
+							Ok(read) => {
+								debug!("jaq: read {read} bytes from {path:?}");
+								Val::Str(buffer.into())
+							}
+							Err(err) => {
+								error!("jaq: failed to read from {path:?}: {err:?}");
+								Val::Null
+							}
+						}
+					}
+					Err(err) => {
+						error!("jaq: failed to open file {path:?}: {err:?}");
+						Val::Null
+					}
+				})))
+			}
+		}),
+	);
+
 	Ok(())
 }