REF: remove deprecated inline linker syntax

Author: chrismacklin

src/GslCore/AstAlgorithms.fs

@@ -94,11 +94,7 @@ let decompile tree =
         | InlineDna({x=dna; positions=_}) -> appendf "/%s/" dna
         | InlineProtein({x=pseq; positions=_}) -> appendf "/$%s/" pseq
         | HetBlock(_) -> append "~"
-        | Gene({x=pg; positions=_}) ->
-            match pg.linker with
-            | Some({l1=l1; l2=l2; orient=o}) ->
-                append (sprintf "%s-%s-%s-%s" l1 l2 o pg.gene)
-            | None -> append pg.gene
+        | Gene({x=gene; positions=_}) -> append gene
         // part mods
         | ParseRelPos({x=rp; positions=_}) ->
             _print rp.i state

src/GslCore/AstExpansion.fs

@@ -461,13 +461,13 @@ let private expandMut
             // Does it contain a mutation modification?
             // Make sure we have only one if so.
             // TODO: this restriction may not be necessary
-            match gp.part.mods |> List.choose modIsMutation with
+            match gp.mods |> List.choose modIsMutation with
             | [] -> p
             | [mutMod] ->
-                if (not (gp.part.gene.[0] = 'G' || gp.part.gene.[0] = 'g')) then
+                if (not (gp.gene.[0] = 'G' || gp.gene.[0] = 'g')) then
                     failwithf
                         "Allele swap gene must be g-type  e.g gABC1$x1234y.  '%c' is not a legal prefix for %s"
-                        (gp.part.gene.[0]) gp.part.gene
+                        (gp.gene.[0]) gp.gene
                 let rg' = getRG a rgs p.pr
 
                 // FIXME: unclear if this was the right behavior, as the rg is selected from both
@@ -492,9 +492,9 @@ let private expandMut
                          | CTERM -> "Cterm"
                          | NONETERM -> "No end preference")
 
-                if not (rg'.IsValid(gp.part.gene.[1..])) then
+                if not (rg'.IsValid(gp.gene.[1..])) then
                     failwithf "Undefined gene '%s' %O\n"
-                        (gp.part.gene.[1..]) (gp.part.where)
+                        (gp.gene.[1..]) (gp.where)
 
                 let asAACheck =
                     match a.pragmas.TryFind("warnoff") with
@@ -520,7 +520,7 @@ let private expandMut
                         verbose
                         rg'
                         codonUsage
-                        gp.part.gene
+                        gp.gene
                         mutMod
                         endPref
                         a.capabilities
@@ -724,10 +724,10 @@ let private expandHB
             // but that's embedded down in the mod list
 
             // Assume they must be using a gene part next to a het block.  Bad?
-            if (not (gp.part.gene.StartsWith("g"))) then
-                failwithf "Heterology block must be adjacent to g part, %s not allowed" gp.part.gene
+            if (not (gp.gene.StartsWith("g"))) then
+                failwithf "Heterology block must be adjacent to g part, %s not allowed" gp.gene
             let s = translateGenePrefix a.pragmas rg' GENE // Start with standard slice
-            let startSlice = applySlices verbose gp.part.mods s // Apply modifiers
+            let startSlice = applySlices verbose gp.mods s // Apply modifiers
             let newSlice =
                 {startSlice with
                     left =
@@ -739,12 +739,12 @@ let private expandHB
             // putting in new consolidated slice.
             let newMods =
                 SLICE(newSlice)
-                ::(gp.part.mods |> List.filter modIsNotSlice)
+                ::(gp.mods |> List.filter modIsNotSlice)
 
             // Prepend backwards as we will flip list at end - watch out, pragmas are reversed as well
             scan
                 a
-                ((GENEPART({gp with part = {gp.part with mods = newMods}}),pr3,fwd3)
+                ((GENEPART({gp with mods = newMods}),pr3,fwd3)
                  ::(INLINEDNA(alt), returnOrFail (pr2.Add("inline")), fwd2)
                  ::(GENEPART(gpUp),pr1,fwd1)
                  ::res)
@@ -778,13 +778,13 @@ let private expandHB
             // but that's embedded down in the mod list
 
             // Assume they must be using a gene part next to a het block.  Bad?
-            if not (gp.part.gene.[0] = 'g') then
+            if not (gp.gene.[0] = 'g') then
                 failwithf
                     "Slices adjacent to het block elements ~ must be gene slices - %s has '%c' gene part"
-                    gp.part.gene gp.part.gene.[0]
+                    gp.gene gp.gene.[0]
 
             let s = translateGenePrefix a.pragmas rg'' GENE // Start with standard slice
-            let startSlice = applySlices verbose gp.part.mods s // Apply modifiers
+            let startSlice = applySlices verbose gp.mods s // Apply modifiers
             let newSlice =
                 {startSlice with
                     left =
@@ -796,12 +796,12 @@ let private expandHB
             // Assemble new mod list by getting rid of existing slice mods and putting in new consolidated slice.
             let newMods =
                 SLICE(newSlice)
-                ::(gp.part.mods |> List.filter modIsNotSlice)
+                ::(gp.mods |> List.filter modIsNotSlice)
             // Prepend backwards as we will flip list at end
             // Note - currently destroy any pragmas attached to the heterology block itself
             scan
                 a
-                ((GENEPART({ gp with part = {gp.part with mods = newMods}} ),pr4,fwd4)
+                ((GENEPART({gp with mods = newMods}), pr4, fwd4)
                  ::(INLINEDNA(newInline), returnOrFail (pr2.Add("inline")),fwd2)
                  ::(GENEPART(gpUp),pr1,fwd1)
                  ::res)
@@ -831,11 +831,11 @@ let private expandHB
             // but that's embedded down in the mod list
 
             // Assume they must be using a gene part next to a het block.  Bad?
-            assert(gp.part.gene.[0] = 'g')
+            assert(gp.gene.[0] = 'g')
 
             // now build up the slice again and apply from scratch to the gene
             let s = translateGenePrefix a.pragmas rg' GENE // Start with standard slice
-            let startSlice = applySlices verbose gp.part.mods s // Apply modifiers
+            let startSlice = applySlices verbose gp.mods s // Apply modifiers
 
             // modify slice to take into account the bit we chopped off
             let newSlice =
@@ -860,14 +860,14 @@ let private expandHB
             // putting in new consolidated slice.
             let newMods =
                 SLICE(newSlice)
-                ::(gp.part.mods |> List.filter modIsNotSlice)
+                ::(gp.mods |> List.filter modIsNotSlice)
             // Prepend backwards as we will flip list at end
             // Note - currently destroy pr2 pragmas associated with the hetblock
             scan
                 a
                 ((GENEPART(gpDown),pr4,fwd4)
                  ::(INLINEDNA(newInline), returnOrFail (pr3.Add("inline")),fwd3)
-                 ::(GENEPART({gp with part = {gp.part with mods = newMods}}),pr1,fwd1)
+                 ::(GENEPART({gp with mods = newMods}), pr1, fwd1)
                  ::res)
                 tl
 

src/GslCore/AstProcess.fs

@@ -810,7 +810,7 @@ let updateDocstringEnvironment = pretransformOnly updateDocstringEnvironmentInne
 let private checkGeneName (rgs:GenomeDefs) (library: Map<string, Dna>) assemblyPragmas node =
     match node with
     | GenePart(pp, gp) ->
-        let geneName = gp.x.gene.[1..].ToUpper()
+        let geneName = gp.x.[1..].ToUpper()
         let partPragmas = getPragmas pp
         getRGNew rgs [partPragmas; assemblyPragmas]
         |> mapMessages (fun s -> errorMessage RefGenomeError s node)

src/GslCore/AstTypes.fs

@@ -141,8 +141,6 @@ type Mutation = {f:char; t:char; loc:int; mType:MType}
 
 type Linker = {l1:string; l2:string; orient:string}
 
-type ParseGene = {gene: string; linker:Linker option}
-
 /// Supported binary operations on nodes.
 type BinaryOperator = | Add | Subtract | Multiply | Divide
 
@@ -201,7 +199,7 @@ and AstNode =
     | InlineDna of Node<string>
     | InlineProtein of Node<string>
     | HetBlock of Node<unit>
-    | Gene of Node<ParseGene>
+    | Gene of Node<string>
     | Assembly of Node<AstNode list>
     // AST nodes for Level 2 syntax support
     | L2Id of Node<L2Id>
@@ -614,11 +612,6 @@ let createPart mods pragmas basePart =
 /// Create a top-level part with empty collections and default values from a base part.
 let createPartWithBase = createPart [] []
 
-/// Create a top-level part given a gene ID.
-let createGenePart (gene: PString) (linker: Linker option) =
-    // The base part for this part will be a Gene AST node.
-    createPartWithBase (Gene({x = {gene = gene.i; linker = linker}; positions = [gene.pos]}))
-
 /// Capture a list of parsed mods and stuff them into their associated part.
 let stuffModsIntoPart astPart mods =
     match astPart with

src/GslCore/Constants.fs

@@ -31,11 +31,6 @@ let defaultRefGenome = "cenpk"
 
 let aaLegal = "ACDEFGHIKLMNPQRSTVWY*" |> Set.ofSeq
 
-/// List of approved linker abbreviations.
-let legalLinkers =
-    [ '0' .. '9' ] @ [ 'A'..'E']
-    |> List.map (fun c -> sprintf "%c" c) |> Set.ofSeq
-
 [<Measure>] type OneOffset
 [<Measure>] type ZeroOffset
 

src/GslCore/DnaCreation.fs

@@ -153,24 +153,24 @@ let getRG (a:Assembly) (rgs:GenomeDefs) (pr:PragmaCollection) =
     | Bad(msgs) -> failwith msgs.[0]
 
 /// Take a genepart and slices and get the actual DNA sequence.
-let realizeSequence verbose  (pragmas:PragmaCollection) fwd (rg:GenomeDef) (gp:GenePartWithLinker) =
+let realizeSequence verbose  (pragmas:PragmaCollection) fwd (rg:GenomeDef) (gp:GenePart) =
 
     if verbose then
         printf "realizeSequence:  fetch fwd=%s %s\n"
-            (if fwd then "y" else "n") gp.part.gene
+            (if fwd then "y" else "n") gp.gene
 
     // Inspect prefix of gene e.g g,t,o,p and see what type of gene part we are starting with
     // Description of part to give in case of error
-    let errorDesc = gp.part.gene
-    let genePart = lookupGenePart errorDesc (gp.part.gene.[0]) (gp.part.mods)
+    let errorDesc = gp.gene
+    let genePart = lookupGenePart errorDesc (gp.gene.[0]) (gp.mods)
 
     // Lookup gene location
-    let feat = rg.get(gp.part.gene.[1..])
+    let feat = rg.get(gp.gene.[1..])
 
     // Come up with an initial slice based on the gene prefix type
     let s = translateGenePrefix pragmas rg genePart
 
-    let finalSlice = applySlices verbose gp.part.mods s
+    let finalSlice = applySlices verbose gp.mods s
     let left = adjustToPhysical feat finalSlice.left
     let right = adjustToPhysical feat finalSlice.right
 
@@ -241,11 +241,7 @@ let expandGenePart
     (a:Assembly)
     specifiedDnaSource
     (ppp:PPP)
-    (gp:GenePartWithLinker) =
-
-    match gp.linker with
-    | None -> () // No linkers were present
-    | Some(l) -> checkLinker l // Test the linkers
+    (gp:GenePart) =
 
     // If the dna source is empty, then we are going to pull the DNA
     // part from the default reference genome, so we should make the
@@ -256,7 +252,7 @@ let expandGenePart
 
     // Check the genes are legal
     //let prefix = gp.part.gene.[0]
-    let g = gp.part.gene.[1..].ToUpper()
+    let g = gp.gene.[1..].ToUpper()
     let rg' = getRG a rgs ppp.pr
 
     if not (rg'.IsValid(g)) then
@@ -268,21 +264,21 @@ let expandGenePart
             // Need to adjust for any slicing carefully since the DNA island is small
             // Validate mods to gene
             let errorRef = match a.name with | None -> sprintf "%A" a | Some(x) -> x
-            validateMods errorRef gp.part.where gp.part.mods
+            validateMods errorRef gp.where gp.mods
 
             // Come up with an initial slice based on the gene prefix type
 
             // Get standard slice range for a gene
             let s = translateGenePrefix a.pragmas rg' GENE
-            let finalSlice = applySlices verbose gp.part.mods s
+            let finalSlice = applySlices verbose gp.mods s
 
             // Ban approx slices to stay sane for now
             if finalSlice.lApprox || finalSlice.rApprox then
                 failwithf
                     "sorry, approximate slices of library genes not supported yet in %A\n"
                     (prettyPrintAssembly a)
 
-            let sliceContext = Library(gp.part.gene)
+            let sliceContext = Library(gp.gene)
 
             let x, y =
                 getBoundsFromSlice finalSlice dna.Length sliceContext
@@ -295,8 +291,8 @@ let expandGenePart
             let orfAnnotation = orfAnnotationFromSlice finalSlice finalDNA.Length ppp.fwd sliceContext
 
             let name1 =
-                if gp.part.mods.Length = 0 then gp.part.gene
-                else (gp.part.gene + (printSlice finalSlice))
+                if gp.mods.Length = 0 then gp.gene
+                else (gp.gene + (printSlice finalSlice))
             let name2 = if ppp.fwd then name1 else "!"+name1
 
             {id = None;
@@ -325,11 +321,11 @@ let expandGenePart
              materializedFrom = Some(ppp);
              annotations = [Orf(orfAnnotation)]}
         else // no :( - wasn't in genome or library
-            failwithf "undefined gene '%s' %O\n" g gp.part.where
+            failwithf "undefined gene '%s' %O\n" g gp.where
     else
         // Inspect prefix of gene e.g g,t,o,p and see what type of gene part we are starting with
-        let errorDesc = gp.part.gene
-        let genePart = lookupGenePart errorDesc (gp.part.gene.[0]) (gp.part.mods)
+        let errorDesc = gp.gene
+        let genePart = lookupGenePart errorDesc gp.gene.[0] gp.mods
         // Lookup gene location
         let rg' = getRG a rgs ppp.pr
 
@@ -354,13 +350,13 @@ let expandGenePart
         // Validate mods to gene
         let errorRef = match a.name with | None -> sprintf "%A" a | Some(x) -> x
 
-        validateMods errorRef gp.part.where gp.part.mods
+        validateMods errorRef gp.where gp.mods
         // Come up with an initial slice based on the gene prefix type
         let s = translateGenePrefix a.pragmas rg' genePart
         if verbose then printf "log: processing %A\n" a
 
         // finalSlice is the consolidated gene relative coordinate of desired piece
-        let finalSlice = applySlices verbose gp.part.mods s
+        let finalSlice = applySlices verbose gp.mods s
 
         // Gene relative coordinates for the gene slice we want
         let finalSliceWithApprox =
@@ -409,7 +405,7 @@ let expandGenePart
         if (left > right && feat.fwd) || (right>left && (not feat.fwd)) then
             failwithf
                 "slice results in negatively lengthed DNA piece for %s\n"
-                (gp.part.gene + (printSlice finalSlice))
+                (gp.gene + (printSlice finalSlice))
 
         /// left' is the genomic coordinate of the start of the element (i.e gene upstream)
         let left', right' = if feat.fwd then left, right else right, left
@@ -431,15 +427,15 @@ let expandGenePart
             |> DnaOps.revCompIf (not ppp.fwd)
 
         let description1 =
-            match gp.part.mods with
-            | [] -> gp.part.gene
+            match gp.mods with
+            | [] -> gp.gene
             | [DOTMOD(d)] ->
                 match d with
-                | "up" -> "u" + gp.part.gene.[1..]
-                | "down" -> "d" + gp.part.gene.[1..]
-                | "mrna" -> "m" + gp.part.gene.[1..]
+                | "up" -> "u" + gp.gene.[1..]
+                | "down" -> "d" + gp.gene.[1..]
+                | "mrna" -> "m" + gp.gene.[1..]
                 | x -> failwithf "unimplemented DOTMOD %s" x
-            | _ -> "g" + gp.part.gene.[1..] + (printSlice finalSlice)
+            | _ -> "g" + gp.gene.[1..] + (printSlice finalSlice)
         let description2 = if ppp.fwd then description1 else "!"+description1
 
         let promStart = {x = -300<OneOffset>; relTo = FivePrime}