Merge pull request #21 from brotherjack/master

Allows for a greater number of datetime strings to be parsed

Manifest.toml

@@ -1,3 +1,5 @@
+# This file is machine-generated - editing it directly is not advised
+
 [[Base64]]
 uuid = "2a0f44e3-6c83-55bd-87e4-b1978d98bd5f"
 
@@ -58,9 +60,14 @@ deps = ["Mmap"]
 uuid = "8bb1440f-4735-579b-a4ab-409b98df4dab"
 
 [[Distributed]]
-deps = ["LinearAlgebra", "Random", "Serialization", "Sockets"]
+deps = ["Random", "Serialization", "Sockets"]
 uuid = "8ba89e20-285c-5b6f-9357-94700520ee1b"
 
+[[ExprTools]]
+git-tree-sha1 = "6f0517056812fd6aa3af23d4b70d5325a2ae4e95"
+uuid = "e2ba6199-217a-4e67-a87a-7c52f15ade04"
+version = "0.1.1"
+
 [[EzXML]]
 deps = ["BinaryProvider", "Libdl", "Pkg", "Printf", "Test"]
 git-tree-sha1 = "ad00b79cca4bb3eabb4209217859c553af4401f5"
@@ -78,7 +85,7 @@ uuid = "0ef565a4-170c-5f04-8de2-149903a85f3d"
 version = "0.5.0"
 
 [[InteractiveUtils]]
-deps = ["LinearAlgebra", "Markdown"]
+deps = ["Markdown"]
 uuid = "b77e0a4c-d291-57a0-90e8-8db25a27a240"
 
 [[IteratorInterfaceExtensions]]
@@ -88,6 +95,7 @@ uuid = "82899510-4779-5014-852e-03e436cf321d"
 version = "0.1.1"
 
 [[LibGit2]]
+deps = ["Printf"]
 uuid = "76f85450-5226-5b5a-8eaa-529ad045b433"
 
 [[Libdl]]
@@ -113,14 +121,20 @@ version = "0.4.0"
 [[Mmap]]
 uuid = "a63ad114-7e13-5084-954f-fe012c677804"
 
+[[Mocking]]
+deps = ["ExprTools"]
+git-tree-sha1 = "916b850daad0d46b8c71f65f719c49957e9513ed"
+uuid = "78c3b35d-d492-501b-9361-3d52fe80e533"
+version = "0.7.1"
+
 [[OrderedCollections]]
 deps = ["Random", "Serialization", "Test"]
 git-tree-sha1 = "c4c13474d23c60d20a67b217f1d7f22a40edf8f1"
 uuid = "bac558e1-5e72-5ebc-8fee-abe8a469f55d"
 version = "1.1.0"
 
 [[Pkg]]
-deps = ["Dates", "LibGit2", "Markdown", "Printf", "REPL", "Random", "SHA", "UUIDs"]
+deps = ["Dates", "LibGit2", "Libdl", "Logging", "Markdown", "Printf", "REPL", "Random", "SHA", "UUIDs"]
 uuid = "44cfe95a-1eb2-52ea-b672-e2afdf69b78f"
 
 [[Printf]]
@@ -215,7 +229,7 @@ uuid = "3bb67fe8-82b1-5028-8e26-92a6c54297fa"
 version = "0.9.4"
 
 [[UUIDs]]
-deps = ["Random"]
+deps = ["Random", "SHA"]
 uuid = "cf7118a7-6976-5b1a-9a39-7adc72f591a4"
 
 [[Unicode]]

Project.toml

@@ -8,4 +8,5 @@ DataFrames = "a93c6f00-e57d-5684-b7b6-d8193f3e46c0"
 Dates = "ade2ca70-3891-5945-98fb-dc099432e06a"
 EzXML = "8f5d6c58-4d21-5cfd-889c-e3ad7ee6a615"
 Geodesy = "0ef565a4-170c-5f04-8de2-149903a85f3d"
+Mocking = "78c3b35d-d492-501b-9361-3d52fe80e533"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"

src/TCX.jl

@@ -1,15 +1,21 @@
 module TCX
-using EzXML, Dates, DataFrames, Geodesy
+using EzXML, Dates, DataFrames, Geodesy, Mocking
 import Base.show
 
 export parse_tcx_dir, parse_tcx_file, getActivityType, getDataFrame, getDistance, getDistance2, getDuration, getAverageSpeed, getAveragePace
 
+const OK = 200
+const CLIENT_ERROR = 400
+const CLIENT_TCX_ERROR = 401
+const NOT_FOUND = 404
+const SERVER_ERROR = 500
+
 struct TrackPoint
     Time::DateTime
     Latitude::Float64
     Longtitude::Float64
     HeartRateBpm::Int32
-    AltitueMeter::Float64
+    AltitudeMeter::Float64
     DistanceMeter::Float64
 end
 
@@ -23,102 +29,127 @@ struct TCXRecord
     TrackPoints::Array{TrackPoint}
 end
 
-DictDateFormats = Dict(
-    :24 => "yyyy-mm-ddTHH:MM:SS.sssZ",
-    :20 => "yyyy-mm-ddTHH:MM:SSZ", 
-    )
-
-function parse_tcx_file(file::String)
-    file_path = abspath(file)
-    if isfile(file_path) == false
-        return 404, nothing
-    end
-
-    xmldoc = try readxml(file_path)
-    catch e
-       if isa(e, EzXML.XMLError)
-           # Not a valid XML document
-           @warn "Invalid XML document: $file_path"
-           return 400, nothing
-       else
-           return 500, nothing
-       end
-    end
-
-    root_element = root(xmldoc)
+function parse_tcx(tcxdoc::EzXML.Document)
+    root_element = root(tcxdoc)
     # Check if TCX
     if nodename(root_element) != "TrainingCenterDatabase"
-        @warn "Invalid TCX document: $file_path"
-        return 400, nothing
+        return CLIENT_TCX_ERROR, nothing
     end
 
     # Type - "/*/*[1]/*[1]/@Sport"
-    aType = nodecontent(findfirst("/*/*[1]/*[1]/@Sport", xmldoc))  
+    aType = nodecontent(findfirst("/*/*[1]/*[1]/@Sport", tcxdoc))
     # Id - "/*/*[1]/*/*[1]"
-    xid =nodecontent(findfirst("/*/*[1]/*/*[1]", xmldoc)) 
-    aId = DateTime(xid,DictDateFormats[length(xid)])
+    xid = nodecontent(findfirst("/*/*[1]/*/*[1]", tcxdoc))
+
+    aId = convertToDateTime(xid)
     # Name = "/*/*[1]/*[1]/*[3]"
-    aName = nodecontent(findfirst("/*/*[1]/*/*[2]", xmldoc))
+    aName = nodecontent(findfirst("/*/*[1]/*/*[2]", tcxdoc))
     # Lap - "/*/*[1]/*/*[2]"
     # TotalSeconds - "/*/*[1]/*/*[2]/*[1]"
-    aTime = parse(Float64, nodecontent(findfirst("/*/*[1]/*/*[2]/*[1]", xmldoc)))
-    aDistance = parse(Float64, nodecontent(findfirst("/*/*[1]/*/*[2]/*[2]", xmldoc)))
+    aTime = parse(Float64, nodecontent(findfirst("/*/*[1]/*/*[2]/*[1]", tcxdoc)))
+    aDistance = parse(Float64, nodecontent(findfirst("/*/*[1]/*/*[2]/*[2]", tcxdoc)))
     # DistanceMeters - "/*/*[1]/*/*[2]/*[2]"
     # AverageHeartRateBpm - "/*/*[1]/*/*[2]/*[5]/*[1]"
-    xbpm = findfirst("/*/*[1]/*/*[2]/*[5]/*[1]", xmldoc) 
+    xbpm = findfirst("/*/*[1]/*/*[2]/*[5]/*[1]", tcxdoc)
     if xbpm === nothing
         aHeartRateBpm = 0
     else
         aHeartRateBpm = parse(Int32, nodecontent(xbpm))
     end
     # TrackPoints - "/*/*[1]/*/*[2]/*[9]/*"
-    tp_Points = findall("/*/*[1]/*/*[2]/*[9]/*", xmldoc)   
+    tp_Points = findall("/*/*[1]/*/*[2]/*[9]/*", tcxdoc)
     aTrackPoints = Array{TrackPoint, size(tp_Points, 1)}[]
     for tp in tp_Points
-        xtime = nodecontent(findfirst("./*[1]", tp))
-        tp_time = DateTime(xtime, DictDateFormats[length(xtime)])
-        xlat = findfirst("./*[2]/*[1]", tp)
-        if xlat !== nothing
-            tp_lat = parse(Float64, nodecontent(xlat))
-        else
-            continue
-        end
-        tp_lont = parse(Float64, nodecontent(findfirst("./*[2]/*[2]", tp)))
-        xbpm = findfirst("./*[5]/*[1]", tp)
-        if xbpm !== nothing
-            tp_bpm = parse(Int32, nodecontent(findfirst("./*[5]/*[1]", tp)))
-        else
-            tp_bm = 0
-        end
-        tp_dist = parse(Float64, nodecontent(findfirst("./*[3]", tp)))
-        tp_alt = parse(Float64, nodecontent(findfirst("./*[4]", tp)))
+        xtime = nodecontent(findfirst("./*[local-name()='Time']", tp))
+        tp_time = convertToDateTime(xtime)
+        tp_lat = parseNode(Float64, "./*[local-name()='Position']/*[local-name()='LatitudeDegrees']", tp)
+        tp_lont = parseNode(Float64, "./*[local-name()='Position']/*[local-name()='LatitudeDegrees']", tp)
+        tp_bpm = parseNode(Int32, "./*[local-name()='HeartRateBpm']/*[1]", tp)
+        tp_dist = parseNode(Float64, "./*[local-name()='TPX']", tp)
+        tp_alt = parseNode(Float64, "./*[local-name()='AltitudeMeters']", tp)
 
         aTrackPoints = vcat(aTrackPoints, TrackPoint(tp_time, tp_lat, tp_lont, tp_bpm, tp_dist, tp_alt))
     end
 
-    return 200, TCXRecord(aId, aName, aType, aDistance, aTime, aHeartRateBpm, aTrackPoints)
+    return OK, TCXRecord(aId, aName, aType, aDistance, aTime, aHeartRateBpm, aTrackPoints)
+end
+
+function parse_tcx_str(str::String)
+    try
+        status, parsed_tcx = parse_tcx(EzXML.parsexml(str))
+        warn_on_tcx_error(status, str, false)
+        return status, parsed_tcx
+    catch e
+        if isa(e, EzXML.XMLError)
+            @error "Invalid XML string: $str"
+            return CLIENT_ERROR, nothing
+        end
+    end
+end
+
+function parse_tcx_file(file::String)
+    file_path = abspath(file)
+    if isfile(file_path) == false
+        return NOT_FOUND, nothing
+    end
+    xmldoc = try @mock EzXML.readxml(file_path)
+    catch e
+       if isa(e, EzXML.XMLError)
+           # Not a valid XML document
+           @warn "Invalid XML document: $file_path"
+           return CLIENT_ERROR, nothing
+       else
+           throw(e)
+       end
+    end
+
+    status, parsed_tcx = parse_tcx(xmldoc)
+    warn_on_tcx_error(status, file_path, true)
+
+    return status, parsed_tcx
+end
+
+function warn_on_tcx_error(status::Int, thing::String, isFile::Bool)
+    if status == CLIENT_TCX_ERROR
+        @warn "Invalid TCX $(isFile ? document : string): $(thing)"
+    end
+end
+
+#=
+= Parses an XML node based on an XPATH and data type.
+=
+= If the node is a `nothing` value, the function returns the data
+= type's version of 0
+=#
+function parseNode(dType, path, node)
+    node_check = findfirst(path, node)
+    if node_check !== nothing
+        return parse(dType, nodecontent(node_check))
+    else
+        return dType(0)
+    end
 end
 
 function parse_tcx_dir(path::String)
     if ispath(path) == false
         @warn "Invalid path: $path"
-        return 500, nothing
+        return SERVER_ERROR, nothing
     end
 
     tcxArray = Array{TCXRecord}[]
     searchdir(path, key) = filter(x->occursin(key, x), readdir(path))
 
     for f in searchdir(path, ".tcx")
         err, tcx = parse_tcx_file(joinpath(path, f))
-        if err == 200
+        if err == OK
             tcxArray = vcat(tcxArray, tcx)
         end
     end
 
     if length(tcxArray) > 0
-        return 200, tcxArray
+        return OK, tcxArray
     else
-        return 404, nothing
+        return NOT_FOUND, nothing
     end
 end
 
@@ -149,8 +180,8 @@ function getDistance2(record::TCXRecord)
     for i in 1:num_of_rows
         if i < num_of_rows
             total_distance += distance(
-                                       LLA(df[i, :Latitude], df[i, :Longtitude], df[i, :AltitueMeter]),
-                                       LLA(df[i+1, :Latitude], df[i+1, :Longtitude], df[i+1, :AltitueMeter])
+                                       LLA(df[i, :Latitude], df[i, :Longtitude], df[i, :AltitudeMeter]),
+                                       LLA(df[i+1, :Latitude], df[i+1, :Longtitude], df[i+1, :AltitudeMeter])
                )
         end
     end
@@ -169,6 +200,34 @@ function getDuration(record::TCXRecord)
     return record.DurationStatic
 end
 
+#=
+= Converts a datetime string into the proper datetime based on string length.
+=
+= Will assume that an ArgumentError is due to
+= https://github.com/JuliaLang/julia/issues/23049 and will attempt to work
+= around this.
+=#
+function convertToDateTime(datestr::String)::DateTime
+    m = match(r"\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}(Z?|\.\d{1,3}Z?)", datestr)
+    format_prefix = "yyyy-mm-ddTHH:MM:SS"
+    if m === nothing
+        msg = "'$(datestr)' is improperly formatted. Must be in the form "
+        msg = msg * "'$(format_prefix)Z' or '$(format_prefix).sssZ'"
+        throw(ArgumentError(msg))
+    else
+        suffix = replace(m.captures[1], r"\d" => "s")
+        try
+            return @mock DateTime(m.match, format_prefix * suffix)
+        catch e
+            if isa(e, ArgumentError)
+                # OK! FINE! NO Z THEN!
+                return DateTime(m.match[1:end-1], format_prefix * (suffix[1:end-1]))
+            else
+                throw(e)
+            end
+        end
+    end
+end
+
 Base.show(io::IO, tcx::TCXRecord) = print(io, "$(tcx.ActivityType) $(tcx.DistanceStatic/1000) km at $(tcx.Id) for $(tcx.DurationStatic) seconds.")
 end #module_end
-

test/runtests.jl

@@ -1,9 +1,13 @@
-using TCX
+using Test, TCX
 
 tcx_sample_file="centry_park_run.tcx"
 tcx_centrypark_run_file="centry_park_run.tcx"
 tcx_treadmill_run_file="treadmill_run.tcx"
 gpx_sample_file="shanghai_marathon_2018.gpx"
 
-include("test_tcx_read_file.jl")
-include("test_tcx_read_dir.jl")
+@testset "TCX tests" begin
+    include("test_tcx_read_file.jl")
+    include("test_tcx_read_dir.jl")
+    include("test_tcx_read_str.jl")
+    include("test_tcx.jl")
+end