schema now local

cli/src/args.rs

@@ -229,6 +229,7 @@ pub struct ZipCrateCommand {
     #[clap(short,long,default_value_t=String::from("./"))]
     pub target_crate: String,
     // Copy and include external reachable data files
+    #[clap(short, long, default_value_t = true)]
     pub external: bool,
 }
 

src/ro_crate/read.rs

@@ -2,8 +2,9 @@
 //! RoCrate data structure
 
 use crate::ro_crate::rocrate::RoCrate;
-use crate::ro_crate::schema::validate_crate_keys;
+use crate::ro_crate::schema::load_rocrate_schema;
 use serde_json;
+use std::collections::HashSet;
 use std::fs;
 use std::io;
 use std::path::{Path, PathBuf};
@@ -80,6 +81,19 @@ pub enum CrateReadError {
     VocabNotValid(String),
 }
 
+impl PartialEq for CrateReadError {
+    fn eq(&self, other: &Self) -> bool {
+        match (self, other) {
+            // We don't compare the actual io::Error or serde_json::Error, just the variant type
+            (CrateReadError::IoError(_), CrateReadError::IoError(_)) => true,
+            (CrateReadError::JsonError(_), CrateReadError::JsonError(_)) => true,
+            // For `VocabNotValid`, we compare the actual error message
+            (CrateReadError::VocabNotValid(a), CrateReadError::VocabNotValid(b)) => a == b,
+            _ => false,
+        }
+    }
+}
+
 impl From<io::Error> for CrateReadError {
     /// Converts an `io::Error` into a `CrateReadError::IoError`.
     fn from(err: io::Error) -> CrateReadError {
@@ -94,6 +108,34 @@ impl From<serde_json::Error> for CrateReadError {
     }
 }
 
+/// Validates that the keys in a given RO-Crate match those defined in the base schema vocabulary.
+///
+/// This function checks the crate's properties against the official RO-Crate context and any embedded vocabularies.
+/// It does not validate properties by dereferencing URIs but rather checks if the properties' keys are recognized.
+pub fn validate_crate_keys(rocrate: &RoCrate) -> bool {
+    match load_rocrate_schema() {
+        Ok(crate_metadata) => {
+            let crate_context: Vec<String> = crate_metadata.context.keys().cloned().collect();
+            let custom_context = rocrate.get_context_items();
+            let vals = RoCrate::get_all_property_values(rocrate);
+
+            // Convert vec1 and vec2 to HashSets for efficient lookup
+            let set1: HashSet<_> = crate_context.into_iter().collect();
+            let set2: HashSet<_> = custom_context.into_iter().collect();
+            let mut invalid_key: Vec<&String> = Vec::new();
+
+            for item in &vals {
+                if !set1.contains(item) && !set2.contains(item) {
+                    invalid_key.push(item);
+                }
+            }
+
+            invalid_key.is_empty()
+        }
+        Err(_e) => false,
+    }
+}
+
 #[cfg(test)]
 mod tests {
     use super::*;
@@ -110,6 +152,25 @@ mod tests {
         assert!(crate_result.is_ok());
     }
 
+    #[test]
+    fn test_read_crate_valid() {
+        let path = fixture_path("_ro-crate-metadata-minimal.json");
+
+        let crate_result = read_crate(&path, true);
+        assert!(crate_result.is_ok());
+    }
+
+    #[test]
+    fn test_read_crate_invalid() {
+        let path = fixture_path("_ro-crate-metadata-broken-schema.json");
+
+        let crate_result = read_crate(&path, true).unwrap_err();
+        match crate_result {
+            CrateReadError::VocabNotValid(_) => (),
+            _ => panic!(),
+        }
+    }
+
     #[test]
     fn test_read_crate_file_not_found() {
         let path = fixture_path("non_existent_file.json");

src/ro_crate/schema.rs

@@ -2,11 +2,9 @@
 
 use crate::ro_crate::constraints::{Id, IdValue, License};
 use crate::ro_crate::rocrate::RoCrate;
-use reqwest;
 use serde::{Deserialize, Serialize};
-use std::collections::{HashMap, HashSet};
-
-const ROCRATE_SCHEMA_1_1: &str = "https://www.researchobject.org/ro-crate/1.1/context.jsonld";
+use std::collections::HashMap;
+const ROCRATE_SCHEMA_1_1: &str = include_str!("../resources/ro_crate_1_1.jsonld");
 
 /// Represents the JSON-LD context of the RO-Crate schema.
 ///
@@ -29,45 +27,48 @@ pub struct RoCrateJSONLDContext {
     pub context: HashMap<String, String>,
 }
 
-/// Downloads the RO-Crate schema from a predefined URL.
+
+/// Loads in RO-Crate schema for validation.
 ///
 /// This function fetches the JSON-LD context defining the RO-Crate schema, attempting to parse it
 /// into an `RoCrateJSONLDContext` struct. It is a synchronous operation and may block the thread
 /// during the request and subsequent parsing.
-pub fn download_rocrate_schema() -> Result<RoCrateJSONLDContext, Box<dyn std::error::Error>> {
-    let res = reqwest::blocking::get(ROCRATE_SCHEMA_1_1)?.text()?;
-    let context = serde_json::from_str(&res)?;
+pub fn load_rocrate_schema() -> Result<RoCrateJSONLDContext, Box<dyn std::error::Error>> {
+    load_rocrate_schema_from_str(ROCRATE_SCHEMA_1_1)
+}
+pub fn load_rocrate_schema_from_str(json_str: &str) -> Result<RoCrateJSONLDContext, Box<dyn std::error::Error>> {
+    let context = serde_json::from_str(json_str)?;
     Ok(context)
 }
 
-/// Validates that the keys in a given RO-Crate match those defined in the base schema vocabulary.
-///
-/// This function checks the crate's properties against the official RO-Crate context and any embedded vocabularies.
-/// It does not validate properties by dereferencing URIs but rather checks if the properties' keys are recognized.
-pub fn validate_crate_keys(rocrate: &RoCrate) -> bool {
-    match download_rocrate_schema() {
-        Ok(crate_metadata) => {
-            let crate_context: Vec<String> = crate_metadata.context.keys().cloned().collect();
-            let custom_context = rocrate.get_context_items();
-            let vals = RoCrate::get_all_property_values(&rocrate);
-
-            // Convert vec1 and vec2 to HashSets for efficient lookup
-            let set1: HashSet<_> = crate_context.into_iter().collect();
-            let set2: HashSet<_> = custom_context.into_iter().collect();
-            let mut invalid_key: Vec<&String> = Vec::new();
+#[cfg(test)]
+mod tests {
+    use super::*;
 
-            for item in &vals {
-                if !set1.contains(item) && !set2.contains(item) {
-                    invalid_key.push(&item);
-                }
-            }
-
-            if invalid_key.is_empty() {
-                true
-            } else {
-                false
+    #[test]
+    fn test_load_rocrate_schema_from_str() {
+        let mock_json = r#"
+        {
+            "@id": "https://w3id.org/ro/crate/1.1/context",
+            "name": ["RO-Crate JSON-LD Context"],
+            "version": "1.1.3",
+            "url": {"@id": "https://w3id.org/ro/crate/1.1"},
+            "schemaVersion": {"@id": "http://schema.org/version/10.0/"},
+            "isBasedOn": [{"@id": "http://schema.org/version/10.0/"}],
+            "license": {"@id": "https://creativecommons.org/publicdomain/zero/1.0/"},
+            "@context": {
+                "3DModel": "http://schema.org/3DModel",
+                "APIReference": "http://schema.org/APIReference"
             }
         }
-        Err(_e) => false,
+        "#;
+
+        let result = load_rocrate_schema_from_str(mock_json);
+        assert!(result.is_ok());
+
+        let context = result.unwrap();
+        assert_eq!(context.version, "1.1.3");
+        assert!(context.context.contains_key("3DModel"));
     }
+
 }

src/ro_crate/write.rs

@@ -136,7 +136,7 @@ pub fn zip_crate(crate_path: &Path, external: bool) -> Result<(), ZipError> {
     {
         let path = entry.path();
 
-        if path == zip_file_name || path == crate_abs {
+        if path == zip_file_name {
             continue;
         }
 

tests/fixtures/_ro-crate-metadata-broken-schema.json

@@ -0,0 +1,28 @@
+
+{ "@context": "https://w3id.org/ro/crate/1.1/context", 
+  "@graph": [
+
+ {
+    "@type": "CreativeWork",
+    "@id": "ro-crate-metadata.json",
+    "conformsTo": {"@id": "https://w3id.org/ro/crate/1.1"},
+    "about": {"@id": "./"}
+ },  
+ {
+    "@id": "./",
+    "identifier": "https://doi.org/10.4225/59/59672c09f4a4b",
+    "@type": "Dataset",
+    "datePublished": "2017",
+    "name": "Data files associated with the manuscript:Effects of facilitated family case conferencing for ...",
+    "description": "Palliative care planning for nursing home residents with advanced dementia ...",
+    "license": {"@id": "https://creativecommons.org/licenses/by-nc-sa/3.0/au/"}
+ },
+ {
+  "@id": "https://creativecommons.org/licenses/by-nc-sa/3.0/au/",
+  "@type": "CreativeWork",
+  "description": "This work is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Australia License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/3.0/au/ or send a letter to Creative Commons, PO Box 1866, Mountain View, CA 94042, USA.",
+  "identifier": "https://creativecommons.org/licenses/by-nc-sa/3.0/au/",
+  "nonschemakey": "Attribution-NonCommercial-ShareAlike 3.0 Australia (CC BY-NC-SA 3.0 AU)"
+ }
+ ]
+}