Revert "parser: sanitize timestamps to RFC3339"

This reverts commit af5e785.

Cargo.toml

@@ -40,7 +40,6 @@ byteorder = "1.4"
 caseless = "0.2"
 chardetng = "0.1"
 chrono = { version = "0.4", features = ["serde"] }
-chrono-tz = { version = "0.8" }
 clap = { version = "3.2", features = ["derive", "env"] }
 colored_json = "3"
 comfy-table = "6.1"

crates/parser/Cargo.toml

@@ -23,7 +23,6 @@ bytes = { workspace = true }
 caseless = { workspace = true }
 chardetng = { workspace = true }
 chrono = { workspace = true }
-chrono-tz = { workspace = true }
 csv = { workspace = true }
 encoding_rs = { workspace = true }
 flate2 = { workspace = true }

crates/parser/src/config/mod.rs

@@ -528,15 +528,11 @@ impl schemars::JsonSchema for ErrorThreshold {
     }
 }
 
-fn default_timezone_string() -> String {
-    "UTC".to_string()
-}
-
 // Fields annotated with `schemars(skip)` will not appear in the JSON schema, and thus won't be
 // shown in the UI. These are things that connectors set programatically when it generates the
 // config. We could consider moving these fields to be CLI arguments if we want a clearer
 // separation.
-#[derive(Clone, Debug, PartialEq, Serialize, Deserialize, schemars::JsonSchema)]
+#[derive(Clone, Debug, Default, PartialEq, Serialize, Deserialize, schemars::JsonSchema)]
 #[schemars(
     title = "Parser Configuration",
     description = "Configures how files are parsed"
@@ -554,11 +550,6 @@ pub struct ParseConfig {
     #[serde(default)]
     pub compression: DefaultNullIsAutomatic<Compression>,
 
-    /// The default timezone to use when parsing timestamps that do not have a timezone. Timezones
-    /// must be specified as a valid IANA name. Defaults to UTC.
-    #[serde(default="default_timezone_string")]
-    pub default_timezone: String,
-
     /// filename is used for format inference. It will be ignored if `format` is specified.
     #[serde(default)]
     #[schemars(skip)]
@@ -591,21 +582,6 @@ pub struct ParseConfig {
     pub content_encoding: Option<String>,
 }
 
-impl Default for ParseConfig {
-    fn default() -> Self {
-        ParseConfig {
-            format: Default::default(),
-            compression: Default::default(),
-            default_timezone: default_timezone_string(),
-            filename: Default::default(),
-            add_record_offset: Default::default(),
-            add_values: Default::default(),
-            content_type: Default::default(),
-            content_encoding: Default::default(),
-        }
-    }
-}
-
 #[derive(Debug, thiserror::Error)]
 pub enum ConfigError {
     #[error("failed to read config: {0}")]
@@ -769,7 +745,7 @@ mod test {
                 }
             },
             "filename": "tha-file",
-            "compression": "zip"
+            "compression": "zip",
         });
 
         let r1: ParseConfig = serde_json::from_value(c1).expect("deserialize config");

crates/parser/src/config/snapshots/parser__config__test__config_schema_is_generated.snap

@@ -40,11 +40,6 @@ expression: schema
         }
       ]
     },
-    "defaultTimezone": {
-      "description": "The default timezone to use when parsing timestamps that do not have a timezone. Timezones must be specified as a valid IANA name. Defaults to UTC.",
-      "default": "UTC",
-      "type": "string"
-    },
     "format": {
       "description": "Determines how to parse the contents. The default, 'Auto', will try to determine the format automatically based on the file extension or MIME type, if available.",
       "default": {

crates/parser/src/format/mod.rs

@@ -2,7 +2,6 @@ pub mod avro;
 pub mod character_separated;
 pub mod json;
 pub mod protobuf;
-pub mod sanitize;
 
 use crate::config::ErrorThreshold;
 use crate::decorate::{AddFieldError, Decorator};
@@ -47,9 +46,6 @@ pub enum ParseError {
 
     #[error("error limit exceeded")]
     ErrorLimitExceeded(ErrorThreshold),
-
-    #[error("failed to sanitize documents: {0}")]
-    SanitizeError(#[from] sanitize::SanitizeError),
 }
 
 /// Runs format inference if the config does not specify a `format`. The expectation is that more
@@ -166,8 +162,7 @@ fn parse_file(
     starting_offset: u64,
 ) -> Result<u64, ParseError> {
     let output = parser.parse(input)?;
-    let sanitized_output = sanitize::sanitize_output(&config, output)?;
-    format_output(&config, sanitized_output, dest, starting_offset)
+    format_output(&config, output, dest, starting_offset)
 }
 
 fn parser_for(format: Format) -> Box<dyn Parser> {