Merge pull request #390 from neelsmith/dev

Add method for `concat_ds` using a `DFParser`

Project.toml

@@ -1,7 +1,7 @@
 name = "Kanones"
 uuid = "107500f9-53d4-4696-8485-0747242ad8bc"
 authors = ["nsmith"]
-version = "0.24.0"
+version = "0.24.1"
 
 [deps]
 AtticGreek = "330c8319-f7ed-461a-8c52-cee5da4c0892"

src/parser/dfparser.jl

@@ -40,4 +40,37 @@ $(SIGNATURES)
 """
 function lexemes(dfp::DFParser)
     Array{String}(DataFrames.select(dfp.df, :Lexeme)) |> unique
+end
+
+
+
+
+
+
+"""Build a new `DFParser` by adding a further dataset
+to an existing parser. 
+
+- `dfp` is an existing `DFParser`.
+- `rulesds` is the dataset used to build `sp`
+- `newdata` is an additional dataset with any new content (rules or vocab)
+"""
+function concat_ds(dfp::DFParser, rulesds::FilesDataset, newdata::FilesDataset; interval = 100)
+    rules_all = vcat(rulesarray(rulesds), rulesarray(newdata))
+    stems_new = stemsarray(newdata)
+
+    analyses = []
+    for (i, stem) in enumerate(stems_new)
+        @info("stem $(i)… $(stem)")
+        if i % interval == 0
+            @info("stem $(i)… $(stem)")
+        end
+        append!(analyses, buildparseable(stem, rules_all, delimiter = ","))
+    end
+    temp_sp = analyses |> StringParser
+    temp_f = tempname()
+    tofile(temp_sp,temp_f; delimiter = ",")
+    dfp2 = dfParser(temp_f)
+    rm(temp_f)
+    DFParser(vcat(dfp.df, dfp2.df) )
+
 end

src/parser/stringparser.jl

@@ -148,8 +148,6 @@ function lexemes(sp::StringParser)
 end
 
 
-#= NO THIS IS ALL WRONG.
-=#
 """Build a new `StringParser` by adding a further dataset
 to an existing parser. 
 

test/assets/tiniest/stems-tables/nouns/onenoun.cex

@@ -0,0 +1,3 @@
+StemUrn|LexicalEntity|Stem|Gender|InflClass|Accent|
+
+nounstems.n72550|lsj.n72550|οἰν|masculine|os_ou|recessive|