From 3e892c24899b6126c2de7a7dc5097d5c56c99837 Mon Sep 17 00:00:00 2001
From: Tom Bland <tom_bland@hotmail.co.uk>
Date: Thu, 20 Mar 2025 13:37:26 +0000
Subject: [PATCH 1/3] Separate functions for allowing/disallowing empty csv
 files

---
 src/input.rs                    | 26 +++++++++++++++++++++++++-
 src/input/agent/search_space.rs |  2 +-
 2 files changed, 26 insertions(+), 2 deletions(-)

diff --git a/src/input.rs b/src/input.rs
index 400bdb19..052b5a24 100644
--- a/src/input.rs
+++ b/src/input.rs
@@ -1,7 +1,7 @@
 //! Common routines for handling input data.
 use crate::agent::AssetPool;
 use crate::model::{Model, ModelFile};
-use anyhow::{ensure, Context, Result};
+use anyhow::{bail, ensure, Context, Result};
 use float_cmp::approx_eq;
 use indexmap::IndexMap;
 use itertools::Itertools;
@@ -26,11 +26,31 @@ pub use time_slice::read_time_slice_info;
 
 /// Read a series of type `T`s from a CSV file.
 ///
+/// Will raise an error if the file is empty.
+///
 /// # Arguments
 ///
 /// * `file_path` - Path to the CSV file
 pub fn read_csv<'a, T: DeserializeOwned + 'a>(
     file_path: &'a Path,
+) -> Result<impl Iterator<Item = T> + 'a> {
+    _read_csv_internal(file_path, true)
+}
+
+/// Read a series of type `T`s from a CSV file.
+///
+/// # Arguments
+///
+/// * `file_path` - Path to the CSV file
+pub fn read_csv_optional<'a, T: DeserializeOwned + 'a>(
+    file_path: &'a Path,
+) -> Result<impl Iterator<Item = T> + 'a> {
+    _read_csv_internal(file_path, false)
+}
+
+fn _read_csv_internal<'a, T: DeserializeOwned + 'a>(
+    file_path: &'a Path,
+    check_empty: bool,
 ) -> Result<impl Iterator<Item = T> + 'a> {
     let vec = csv::Reader::from_path(file_path)
         .with_context(|| input_err_msg(file_path))?
@@ -38,6 +58,10 @@ pub fn read_csv<'a, T: DeserializeOwned + 'a>(
         .process_results(|iter| iter.collect_vec())
         .with_context(|| input_err_msg(file_path))?;
 
+    if check_empty && vec.is_empty() {
+        bail!("CSV file {} cannot be empty", file_path.display());
+    }
+
     Ok(vec.into_iter())
 }
 
diff --git a/src/input/agent/search_space.rs b/src/input/agent/search_space.rs
index 0f19f805..ad1ac477 100644
--- a/src/input/agent/search_space.rs
+++ b/src/input/agent/search_space.rs
@@ -81,7 +81,7 @@ pub fn read_agent_search_space(
     milestone_years: &[u32],
 ) -> Result<HashMap<Rc<str>, Vec<AgentSearchSpace>>> {
     let file_path = model_dir.join(AGENT_SEARCH_SPACE_FILE_NAME);
-    let iter = read_csv::<AgentSearchSpaceRaw>(&file_path)?;
+    let iter = read_csv_optional::<AgentSearchSpaceRaw>(&file_path)?;
     read_agent_search_space_from_iter(iter, agents, process_ids, commodities, milestone_years)
         .with_context(|| input_err_msg(&file_path))
 }

From ba94748efda08baf24262a972446159093a0db55 Mon Sep 17 00:00:00 2001
From: Tom Bland <tom_bland@hotmail.co.uk>
Date: Thu, 20 Mar 2025 13:41:56 +0000
Subject: [PATCH 2/3] Add test

---
 src/input.rs | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/src/input.rs b/src/input.rs
index 052b5a24..8bbef276 100644
--- a/src/input.rs
+++ b/src/input.rs
@@ -309,6 +309,14 @@ mod tests {
                 }
             ]
         );
+
+        // File with no data (only column headers)
+        let file_path = create_csv_file(dir.path(), "id,value\n");
+        assert!(read_csv::<Record>(&file_path).is_err());
+        assert!(read_csv_optional::<Record>(&file_path)
+            .unwrap()
+            .next()
+            .is_none());
     }
 
     #[test]

From 7d7372fe408628bddc8c0350d43559e59c55e499 Mon Sep 17 00:00:00 2001
From: Tom Bland <tom_bland@hotmail.co.uk>
Date: Fri, 21 Mar 2025 08:26:21 +0000
Subject: [PATCH 3/3] Move code for checking `Vec`

---
 src/input.rs | 20 +++++++++-----------
 1 file changed, 9 insertions(+), 11 deletions(-)

diff --git a/src/input.rs b/src/input.rs
index 8bbef276..0e185b20 100644
--- a/src/input.rs
+++ b/src/input.rs
@@ -34,7 +34,11 @@ pub use time_slice::read_time_slice_info;
 pub fn read_csv<'a, T: DeserializeOwned + 'a>(
     file_path: &'a Path,
 ) -> Result<impl Iterator<Item = T> + 'a> {
-    _read_csv_internal(file_path, true)
+    let vec = _read_csv_internal(file_path)?;
+    if vec.is_empty() {
+        bail!("CSV file {} cannot be empty", file_path.display());
+    }
+    Ok(vec.into_iter())
 }
 
 /// Read a series of type `T`s from a CSV file.
@@ -45,24 +49,18 @@ pub fn read_csv<'a, T: DeserializeOwned + 'a>(
 pub fn read_csv_optional<'a, T: DeserializeOwned + 'a>(
     file_path: &'a Path,
 ) -> Result<impl Iterator<Item = T> + 'a> {
-    _read_csv_internal(file_path, false)
+    let vec = _read_csv_internal(file_path)?;
+    Ok(vec.into_iter())
 }
 
-fn _read_csv_internal<'a, T: DeserializeOwned + 'a>(
-    file_path: &'a Path,
-    check_empty: bool,
-) -> Result<impl Iterator<Item = T> + 'a> {
+fn _read_csv_internal<'a, T: DeserializeOwned + 'a>(file_path: &'a Path) -> Result<Vec<T>> {
     let vec = csv::Reader::from_path(file_path)
         .with_context(|| input_err_msg(file_path))?
         .into_deserialize()
         .process_results(|iter| iter.collect_vec())
         .with_context(|| input_err_msg(file_path))?;
 
-    if check_empty && vec.is_empty() {
-        bail!("CSV file {} cannot be empty", file_path.display());
-    }
-
-    Ok(vec.into_iter())
+    Ok(vec)
 }
 
 /// Parse a TOML file at the specified path.