Merge pull request #20 from rtk-rs/sp3-formatting

- SP3 formatting
   - add clock and prediction flags
   - add maneuver flags
   - add velocities and derivatives
- Introduce Production Attributes and filename conventions
- raised an issue with MJD fract

Author: gwbres

README.md

@@ -83,6 +83,23 @@ for (epoch, sv, (x_km_ecef, y_km_ecef, z_km_ecef)) in sp3.satellites_position_km
 for (epoch, sv, clock) in sp3.satellites_clock_offset_sec_iter() {
 
 }
+
+// Dump to file
+sp3.to_file("/tmp/test.txt")
+    .unwrap();
+
+let attributes = sp3.prod_attributes
+    .expect("exists for files that follow standards conventions!");
+
+// "ESA0OPSRAP_20232390000_01D_15M_ORB.SP3.gz
+assert_eq!(attributes.agency, "ESA");
+assert_eq!(attributes.batch_id, 0);
+assert_eq!(attributes.availability, Availability::Rapid);
+assert_eq!(attributes.release_date.year, 2023);
+assert_eq!(attributes.release_date.doy, 239);
+assert_eq!(attributes.release_period, ReleasePeriod::Daily);
+assert_eq!(attributes.sampling_period, sp3.header.sampling_period);
+assert!(attributes.gzip_compressed);
 ```
 
 ## Lib features
@@ -108,14 +125,10 @@ The `Nyx-Space/ANISE` feature is the heaviest library option.
 
 ## Satellite attitude interpolation
 
-Satellite (SV) attitude interpolation is a major topic in SP3 processing. 
-Because quite often, you will have to match data provided by SP3 (high precision) other
-datasets, oftentimes expressed in different timescales and most often sampled at higher rate.
-
-Indeed, SP3 is published by laboratories with a typical fit rate of 15'. 
-
-To answer the requirement of SP3 processing inside a broader geodetic processing pipeline,
-this library proposes a few sets of method
+Satellite (SV) attitude interpolation is a major topic in SP3 processing.  
+Typically, SP3 data has to be matched (in time) with other data which requires interpolation.
+In order to preserve the quality of the SP3 fit, it is recommended to use a high order Lagrangian
+interpolation. This library provides SP3 interpolation as is, because it does not involve external libraries:
 
 - `[SP3.satellite_position_interp()]` will design the interpolation kernel
 to which you can apply your custom interpolation function
@@ -130,10 +143,8 @@ method with a order of 11, which is typically used to preserve SP3 precision
 method with a order of 17, which is way more than enough and should be used in processing
 pipelines where processing speed and resource consumption is not an issue. 
 
-Note that SP3 provides attitude with 10⁻³m precision.
-
-The (timewise) interpolation kernel is only feasible for odd interpolation order (at the moment),
-for simplicity. The extracted kernel is therefore:
+:warning: our interpolation method does not support even interpolation orders.
+The extracted kernel is therefore:
 
 - `tmin = (N +1)/2 * τ`
 - `tmax =  T(n-1) - (N +1)/2 * τ`
@@ -210,6 +221,34 @@ let sp3 = sp3_a.merge(&sp3_b);
 assert!(sp3.is_ok());
 ```
 
+## QC: Timescale Transposition
+
+Use the `Timeshift` trait to transpose the SP3 into other timescales. Coarse
+and precise transpositions are both supported. The precise transposition method
+requires a correction database.
+
+```rust
+use sp3::prelude::*;
+use std::path::PathBuf;
+
+let path = PathBuf::new()
+    .join(env!("CARGO_MANIFEST_DIR"))
+    .join("data/SP3")
+    .join("C")
+    .join("ESA0OPSRAP_20232390000_01D_15M_ORB.SP3.gz");
+
+// Typical GPST SP3
+let gpst_sp3 = SP3::from_gzip_file(&path)
+    .unwrap();
+
+// // Transpose to GST
+// let gst_sp3 = gpst_sp3.timeshift(TimeScale::GST);
+
+// Dump as GST file
+// gst_sp3.to_file("/tmp/gst.txt")
+//    .unwrap();
+```
+
 ## License
 
 This library is part of the [RTK-rs framework](https://github.com/rtk-rs) which