Merge branch 'master' into dependabot/github_actions/codecov/codecov-…

…action-5

.gitignore

@@ -2,6 +2,7 @@
 *.jl.*.cov
 *.jl.mem
 Manifest.toml
+Manifest-v*.toml
 docs/build/
 docs/site/
 .DS_Store

OndaEDFSchemas.jl/Project.toml

@@ -1,7 +1,7 @@
 name = "OndaEDFSchemas"
 uuid = "9c87d999-769b-4741-85b2-6f554d09e731"
 authors = ["Beacon Biosignals, Inc."]
-version = "0.2.2"
+version = "0.2.3"
 
 [deps]
 Legolas = "741b9549-f6ed-4911-9fbf-4a1c0c97f0cd"

OndaEDFSchemas.jl/src/OndaEDFSchemas.jl

@@ -7,7 +7,7 @@ using Onda: LPCM_SAMPLE_TYPE_UNION, onda_sample_type_from_julia_type,
             AnnotationV1
 using UUIDs
 
-export PlanV1, PlanV2, FilePlanV1, FilePlanV2, EDFAnnotationV1
+export PlanV1, PlanV2, PlanV3, FilePlanV1, FilePlanV2, FilePlanV3, EDFAnnotationV1
 
 @schema "ondaedf.plan" Plan
 
@@ -65,7 +65,33 @@ end
     error::Union{Nothing,String} = coalesce(error, nothing)
 end
 
-
+@version PlanV3 begin
+    # EDF.SignalHeader fields
+    label::String
+    transducer_type::String
+    physical_dimension::String
+    physical_minimum::Float32
+    physical_maximum::Float32
+    digital_minimum::Float32
+    digital_maximum::Float32
+    prefilter::String
+    samples_per_record::Int32
+    # EDF.FileHeader field
+    seconds_per_record::Float64
+    # Onda.SignalV2 fields (channels -> channel), may be missing
+    recording::Union{UUID,Missing} = lift(UUID, recording)
+    sensor_type::Union{Missing,AbstractString} = lift(_validate_signal_sensor_type, sensor_type)
+    sensor_label::Union{Missing,AbstractString} = lift(_validate_signal_sensor_label,
+                                                       coalesce(sensor_label, sensor_type))
+    channel::Union{Missing,AbstractString} = lift(_validate_signal_channel, channel)
+    sample_unit::Union{Missing,AbstractString} = lift(String, sample_unit)
+    sample_resolution_in_unit::Union{Missing,Float64}
+    sample_offset_in_unit::Union{Missing,Float64}
+    sample_type::Union{Missing,AbstractString} = lift(onda_sample_type_from_julia_type, sample_type)
+    sample_rate::Union{Missing,Float64}
+    # errors, use `nothing` to indicate no error
+    error::Union{Nothing,String} = coalesce(error, nothing)
+end
 
 const PLAN_DOC_TEMPLATE = """
     @version PlanV{{ VERSION }} begin
@@ -78,7 +104,7 @@ const PLAN_DOC_TEMPLATE = """
         digital_minimum::Float32
         digital_maximum::Float32
         prefilter::String
-        samples_per_record::Int16
+        samples_per_record::{{ SAMPLES_PER_RECORD_TYPE }}
         # EDF.FileHeader field
         seconds_per_record::Float64
         # Onda.SignalV{{ VERSION }} fields (channels -> channel), may be missing
@@ -108,19 +134,22 @@ conversion.  The columns are the union of
 function _plan_doc(v)
     uniques = if v == 1
         ["kind::Union{Missing,AbstractString}"]
-    elseif v == 2
+    elseif v == 2 || v == 3
         ["sensor_type::Union{Missing,AbstractString}",
          "sensor_label::Union{Missing,AbstractString}"]
     else
         throw(ArgumentError("Invalid version"))
     end
+    samples_per_record_type = v in (1,2) ? "Int16" : "Int32"
     unique_lines = join(map(s -> "        $s", uniques), "\n")
     s = replace(PLAN_DOC_TEMPLATE, "{{ VERSION }}" => v)
+    s = replace(s, "{{ SAMPLES_PER_RECORD_TYPE }}" => samples_per_record_type)
     return replace(s, "{{ SAMPLES_INFO_UNIQUE_FIELDS }}" => unique_lines)
 end
 
 @doc _plan_doc(1) PlanV1
 @doc _plan_doc(2) PlanV2
+@doc _plan_doc(3) PlanV3
 
 @schema "ondaedf.file-plan" FilePlan
 
@@ -134,6 +163,11 @@ end
     onda_signal_index::Int
 end
 
+@version FilePlanV3 > PlanV3 begin
+    edf_signal_index::Int
+    onda_signal_index::Int
+end
+
 const FILE_PLAN_DOC_TEMPLATE = """
     @version FilePlanV{{ VERSION }} > PlanV{{ VERSION }} begin
         edf_signal_index::Int
@@ -158,8 +192,11 @@ end
 
 @doc _file_plan_doc(1) FilePlanV1
 @doc _file_plan_doc(2) FilePlanV2
+@doc _file_plan_doc(3) FilePlanV3
 
-const OndaEDFSchemaVersions = Union{PlanV1SchemaVersion,PlanV2SchemaVersion,FilePlanV1SchemaVersion,FilePlanV2SchemaVersion}
+const OndaEDFSchemaVersions = Union{PlanV1SchemaVersion,FilePlanV1SchemaVersion,
+                                    PlanV2SchemaVersion,FilePlanV2SchemaVersion,
+                                    PlanV3SchemaVersion,FilePlanV3SchemaVersion}
 Legolas.accepted_field_type(::OndaEDFSchemaVersions, ::Type{String}) = AbstractString
 # we need this because Arrow write can introduce a Missing for the error column
 # (I think because of how missing/nothing sentinels are handled?)

OndaEDFSchemas.jl/test/runtests.jl

@@ -19,14 +19,20 @@ function mock_plan(; v, rng=GLOBAL_RNG)
     ingested = rand(rng, Bool)
     specific_kwargs = if v == 1
         (; kind=ingested ? "eeg" : missing)
-    elseif v == 2
+    elseif v in (2, 3)
         (; sensor_type=ingested ? "eeg" : missing,
          sensor_label=ingested ? "eeg" : missing)
     else
         error("Invalid version")
     end
     errored = !ingested && rand(rng, Bool)
-    PlanVersion = v == 1 ? PlanV1 : PlanV2
+    PlanVersion = if v == 1
+        PlanV1
+    elseif v == 2
+        PlanV2
+    else
+        PlanV3
+    end
     return PlanVersion(; label="EEG CZ-M1",
                        transducer_type="Ag-Cl electrode",
                        physical_dimension="uV",
@@ -55,15 +61,21 @@ end
 
 function mock_file_plan(; v, rng=GLOBAL_RNG)
     plan = mock_plan(; v, rng)
-    PlanVersion = v == 1 ? FilePlanV1 : FilePlanV2
+    PlanVersion = if v == 1
+        FilePlanV1
+    elseif v == 2
+        FilePlanV2
+    else
+        FilePlanV3
+    end
     return PlanVersion(Tables.rowmerge(plan;
                                        edf_signal_index=rand(rng, Int),
                                        onda_signal_index=rand(rng, Int)))
 end
 
-@testset "Schema version $v" for v in (1, 2)
+@testset "Schema version $v" for v in (1, 2, 3)
     SamplesInfo = v == 1 ? Onda.SamplesInfoV1 : SamplesInfoV2
-    
+
     @testset "ondaedf.plan@$v" begin
         rng = StableRNG(10)
         plans = mock_plan(30; v, rng)

Project.toml

@@ -1,7 +1,7 @@
 name = "OndaEDF"
 uuid = "e3ed2cd1-99bf-415e-bb8f-38f4b42a544e"
 authors = ["Beacon Biosignals, Inc."]
-version = "0.12.4"
+version = "0.13.0"
 
 [deps]
 Compat = "34da2185-b29b-5c13-b0c7-acf172513d20"
@@ -18,12 +18,13 @@ UUIDs = "cf7118a7-6976-5b1a-9a39-7adc72f591a4"
 
 [compat]
 Compat = "3.32, 4"
-EDF = "0.7"
+EDF = "0.8"
 FilePathsBase = "0.9"
 Legolas = "0.5"
 Onda = "0.15"
-OndaEDFSchemas = "0.2.1"
+OndaEDFSchemas = "0.2.3"
 PrettyTables = "1.3, 2"
+SparseArrays = "1"
 StableRNGs = "1"
 StatsBase = "0.33, 0.34"
 Tables = "1.4"
@@ -33,9 +34,10 @@ julia = "1.6"
 [extras]
 FilePathsBase = "48062228-2e41-5def-b9a4-89aafe57970f"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
+SparseArrays = "2f01184e-e22b-5df5-ae63-d93ebab69eaf"
 StableRNGs = "860ef19b-820b-49d6-a774-d7a799459cd3"
 Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 
 [targets]
-test = ["FilePathsBase", "Test", "Random", "StableRNGs", "Statistics"]
+test = ["FilePathsBase", "Test", "Random", "SparseArrays", "StableRNGs", "Statistics"]

docs/Project.toml

@@ -2,4 +2,4 @@
 Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
 
 [compat]
-Documenter = "0.26"
+Documenter = "1"

docs/src/api.md

@@ -9,7 +9,7 @@ CurrentModule = OndaEDF
 OndaEDF.jl prefers "self-service" import over "automagic", and provides
 functionality to extract
 [`Onda.Samples`](https://beacon-biosignals.github.io/Onda.jl/stable/#Samples-1)
-and [`EDFAnnotationV1`](@ref)s (which extend 
+and [`EDFAnnotationV1`](@ref)s (which extend
 [`Onda.AnnotationV1`](https://beacon-biosignals.github.io/Onda.jl/stable/#Onda.AnnotationV1)s)
 from an `EDF.File`.  These can be written to disk (with
 [`Onda.store`](https://beacon-biosignals.github.io/Onda.jl/stable/#Onda.store) /
@@ -33,9 +33,17 @@ EDFAnnotationV1
 
 ### Import plan table schemas
 
+!!! note "Plan version is dependent on EDF.jl version"
+    The utilized plan version is dependent on the EDF.jl version.
+    For EDF.jl 0.8+, V3 are used, while V2 is used for for EDF.jl 0.7.
+    The change from V2 to V3 reflects the change from [`Int16` to `Int32`
+    in EDF.jl's `samples_per_record`](https://github.com/beacon-biosignals/EDF.jl/releases/tag/v0.8.0).
+
 ```@docs
 PlanV2
+PlanV3
 FilePlanV2
+FilePlanV3
 write_plan
 ```
 
@@ -63,6 +71,12 @@ OndaEDF.promote_encodings
 onda_to_edf
 ```
 
+#### Internal export utilities
+
+```@docs
+OndaEDF.reencode_samples
+```
+
 ## Deprecations
 
 To support deserializing plan tables generated with old versions of OndaEDF +

docs/src/convert-to-onda.md

@@ -101,7 +101,7 @@ To break this label specification down:
 - `["emg", "chin"]`: indicates that this specification is for `sensor_type` of `"emg"`, for which `"chin"` is accepted as a non-canonical alternative.
 - `["chin" => ["emg"]]` specifies that there's _one_ possible channel for this sensor type, whose canonical label (used in the output) is `"chin"`, but for which an alternative of `"emg"` is accepted.
 
-This essentially tricks OndaEDF into treating `CHIN EMG` as sensor type of "`CHIN` (alternative for sensor type of `EMG`), and channel label "`EMG` (alternative way to specify `CHIN`)".  For more details on how this matching is carried out, see [`plan_edf_to_onda_samples`](@ref) and [`match_labels`](@ref).
+This essentially tricks OndaEDF into treating `CHIN EMG` as sensor type of "`CHIN` (alternative for sensor type of `EMG`), and channel label "`EMG` (alternative way to specify `CHIN`)".  For more details on how this matching is carried out, see [`plan_edf_to_onda_samples`](@ref) and the internal [`OndaEDF.match_edf_label`](@ref).
 
 #### Preprocess signal headers
 
@@ -117,8 +117,8 @@ edf = EDF.File(my_edf_file_path)
 
 function corrected_header(signal::EDF.Signal)
     header = signal.header
-    return Tables.rowmerge(header; 
-                           label=header.transducer_type, 
+    return Tables.rowmerge(header;
+                           label=header.transducer_type,
                            transducer_type=header.label)
 end
 
@@ -200,7 +200,7 @@ If any errors _were_ encountered, you may need to iterate further.
 
 The final step is to store both the `Onda.Samples` and the executed plan in some persistent storage.
 For storing `Onda.Samples`, see [`Onda.store`](https://beacon-biosignals.github.io/Onda.jl/stable/#Onda.store), which supports serializing LPCM-encoded samples to [any "path-like" type](https://beacon-biosignals.github.io/Onda.jl/stable/#Support-For-Generic-Path-Like-Types) (i.e., anything that provides a method for `write`).
-For storing the plan, use [`OndaEDF.write_plan`](@ref) (or `Legolas.write(file_path, plan, FilePlanV2SchemaVersion())` (see the documentation for [`Legolas.write`](https://beacon-biosignals.github.io/Legolas.jl/stable/#Legolas.write) and [`FilePlanV2`](@ref).
+For storing the plan, use [`OndaEDF.write_plan`](@ref) (or `Legolas.write(file_path, plan, FilePlanV3SchemaVersion())` (see the documentation for [`Legolas.write`](https://beacon-biosignals.github.io/Legolas.jl/stable/#Legolas.write) and [`FilePlanV3`](@ref OndaEDF.FilePlanV3).
 
 ## Batch conversion of many EDFs
 
@@ -262,7 +262,7 @@ function plan_all(edf_paths, files; kwargs...)
         plan = DataFrame(plan)
         # make sure this is the same every time this function is re-run!
         recording = uuid5(NAMESPACE, string(origin_uri))
-        return insertcols!(plan, 
+        return insertcols!(plan,
                            :origin_uri => origin_uri,
                            :recording => recording)
     end

src/OndaEDF.jl

@@ -29,7 +29,7 @@ Write a plan table to `io_or_path` using `Legolas.write`, using the
 """
 function write_plan(io_or_path, plan_table; kwargs...)
     return Legolas.write(io_or_path, plan_table,
-                         Legolas.SchemaVersion("ondaedf.file-plan", 2);
+                         Legolas.SchemaVersion("ondaedf.file-plan", 3);
                          kwargs...)
 end
 

src/export_edf.jl

@@ -23,7 +23,7 @@ end
 const DATA_RECORD_SIZE_LIMIT = 30720
 const EDF_BYTE_LIMIT = 8
 
-edf_sample_count_per_record(samples::Samples, seconds_per_record::Float64) = Int16(samples.info.sample_rate * seconds_per_record)
+edf_sample_count_per_record(samples::Samples, seconds_per_record::Float64) = Int32(samples.info.sample_rate * seconds_per_record)
 
 _rationalize(x) = rationalize(x)
 _rationalize(x::Int) = x // 1