diff --git a/src/call.py b/src/call.py
index 53bb68015bbb2413a558c8abe88b92f31a58d329..bdc880c4eab8ece4c070e938433dc29e031c92ae 100644
--- a/src/call.py
+++ b/src/call.py
@@ -26,7 +26,6 @@ class Call(Executor):
def __str__(self):
return f"Call_{id(self)}"
-
def get_code(self, clean_pipe = False, get_OG=False):
if(get_OG):
@@ -47,15 +46,34 @@ class Call(Executor):
code = f"{tag_to_add}\n{code}"
index = 1
for param in self.parameters:
+ param_new_name = f"{self.get_first_element_called().get_name()}_param_{index}"
+
#Case the param is a call
if(param.get_type()=="Call"):
- param_new_name = f"{self.get_first_element_called().get_name()}_param_{index}"
- code = code.replace(param.get_code(), param_new_name)
+ code = code.replace(param.get_code(get_OG=True), param_new_name)
new_bit = param.simplify_code(new_name = param_new_name)
code = code.replace(tag_to_add, f"{tag_to_add}\n{new_bit}")
-
+
#Case the param is an operation
- #TODO
+ elif(param.get_type()=="Operation"):
+
+ code = code.replace(param.get_code(get_OG=True), param_new_name)
+ lines = param.simplify_code().split('\n')
+ if(len(lines)==1):
+ new_bit = f"{param_new_name} = {lines[0]}"
+ else:
+ head = '\n'.join(lines[:-1])
+ new_bit = f"{head}\n{param_new_name} = {lines[-1]}"
+ code = code.replace(tag_to_add, f"{tag_to_add}\n{new_bit}")
+
+ #Case Channel
+ elif(param.get_type()=="Channel"):
+ None
+ elif(param.get_type()=="Emmited"):
+ None
+ else:
+ print(param.get_code(), param.get_type())
+ raise Exception("This shouldn't happen")
index+=1
return code.replace(tag_to_add, "").strip()
diff --git a/src/graph.py b/src/graph.py
index bcc826f6a78e803675ef0cf6e3e0709d98824018..084b061185e3c163ba313360dc1d002d9a42aa6e 100644
--- a/src/graph.py
+++ b/src/graph.py
@@ -300,8 +300,8 @@ class Graph():
def get_user_view_graph(self, relevant_processes = []):
#For now i'm only gonna work from the flattened dico
- self.initialise_flattened_dico(self.dico_process_dependency_graph)
- #self.initialise_flattened_dico(self.full_dico)
+ #self.initialise_flattened_dico(self.dico_process_dependency_graph)
+ self.initialise_flattened_dico(self.full_dico)
dico = self.dico_flattened
user_view, self.new_nodes_user_view = relev_user_view_builder(dico, relevant_modules=relevant_processes)
diff --git a/src/main_DSL2.py b/src/main_DSL2.py
index cc4a188df1d1f1bf1ae7c4e895cf963c234a40b7..ab27fc75cfd879654424e5a912db873a5758ae12 100644
--- a/src/main_DSL2.py
+++ b/src/main_DSL2.py
@@ -15,12 +15,26 @@ class Main_DSL2(Nextflow_Building_Blocks):
self.initialised = False
self.conditions=None
- def get_all_executors(self):
- tab = []
- tab+=self.get_executors()
- for sub in self.get_subworkflows_called():
- tab+=sub.get_executors()
- return tab
+ def get_all_executors(self, dico):
+ for e in self.get_executors():
+ dico[e] = 1
+
+ for exe in self.get_executors():
+ if(exe.get_type()=="Call"):
+ first = exe.get_first_element_called()
+ if(first.get_type()=="Subworkflow"):
+ first.get_all_executors(dico)
+ elif(exe.get_type()=="Operation"):
+ for o in exe.get_origins():
+ if(o.get_type()=="Call"):
+ first = o.get_first_element_called()
+ if(first.get_type()=="Subworkflow"):
+ print("here")
+ first.get_all_executors(dico)
+ else:
+ raise Exception("This shouldn't happen")
+
+
def get_channels(self):
return self.channels
diff --git a/src/nextflow_building_blocks.py b/src/nextflow_building_blocks.py
index f54182b57d4b83f94f7bf10e325bcb468e25564c..82a782269820625db1d8e5db1f37196df6c60c0d 100644
--- a/src/nextflow_building_blocks.py
+++ b/src/nextflow_building_blocks.py
@@ -16,7 +16,7 @@ class Nextflow_Building_Blocks:
self.processes = []
self.channels = []
- self.DSL = ""
+ #self.DSL = ""
#DSL2
self.includes = []
self.main = None
@@ -24,7 +24,7 @@ class Nextflow_Building_Blocks:
self.subworkflows = []
self.functions=[]
-
+
#---------------------------------
#AUXILIARY METHODS FOR ALL CLASSES
diff --git a/src/nextflow_file.py b/src/nextflow_file.py
index 3a3054d5e713e91e777fd3d7ae326112fd1b5e43..e25e2f03eca47db88dfaaf403093c38906c30b0c 100644
--- a/src/nextflow_file.py
+++ b/src/nextflow_file.py
@@ -44,8 +44,14 @@ class Nextflow_File(Nextflow_Building_Blocks):
#self.extract_metadata()
self.check_file_correctness_after_DSL()
self.set_null()
+
+ def set_new_code(self, code):
+ #self.DSL = self.get_DSL()
+ Nextflow_Building_Blocks.__init__(self, code)
-
+ def get_address(self):
+ return self.get_file_address()
+
def get_name_file(self):
name = self.get_file_address().split('/')[-1]
return name[:-3]
@@ -133,8 +139,8 @@ class Nextflow_File(Nextflow_Building_Blocks):
self.all_includes = []
self.added_2_rocrate = False
- def get_all_executors(self):
- return self.main.get_all_executors()
+ def get_all_executors(self, dico):
+ return self.main.get_all_executors(dico)
def extract_metadata(self):
@@ -546,7 +552,7 @@ class Nextflow_File(Nextflow_Building_Blocks):
self.extract_functions()
#Extract Executors
- self.extract_executors()
+ #self.extract_executors()
#Analyse Executors
for e in self.executors:
diff --git a/src/operation.py b/src/operation.py
index 25fa23e2bbc7c4397051cb49309025a478217d6d..ee7964968fa29b0e67c477a81da6fd771d339836 100644
--- a/src/operation.py
+++ b/src/operation.py
@@ -875,6 +875,34 @@ class Operation(Executor):
def convert_to_DSL2(self):
code = self.get_code(get_OG=True)
return operation_2_DSL2(code, self)
+
+ #Method that rewrites operations to simplify it -> decompose it into multiple line -> to be able to manipulate the calls in a easier way
+ def simplify_code(self):
+ code = self.get_code(get_OG=True)
+ index = 1
+ operation_id = str(self)[-7:-2]
+
+ def add_origin_equals(call, index):
+ simplified_code = call.simplify_code()
+ lines = simplified_code.split('\n')
+ return f"{simplified_code}\noperation_{operation_id}_{index} = {call.get_first_element_called().get_name()}.out[0]"
+ #if(len(lines)==1):
+ # return f"operation_{operation_id}_{index} = {simplified_code}"
+ #else:
+ # head = '\n'.join(lines[:-1])
+ # return f"{head}\noperation_{operation_id}_{index} = {lines[-1]}"
+
+ to_add = []
+ for o in self.origins:
+ if(o.get_type()=="Call"):
+ to_add.append(add_origin_equals(o, index))
+ code = code.replace(o.get_code(get_OG=True), f"operation_{operation_id}_{index}")
+ index += 1
+
+ to_add.reverse()
+ for c in to_add:
+ code = f"{c}\n{code}"
+ return code
diff --git a/src/outils.py b/src/outils.py
index e06bda1df0f21c01f57a62a95a2626efd0175e64..e33d6a0e1fd79c05c5d3fcd4d53f7d39cc812e02 100644
--- a/src/outils.py
+++ b/src/outils.py
@@ -1105,3 +1105,65 @@ def operation_2_DSL2(code, origin):
code+=f"\n{body}.set{{{gives.get_code()}}}"
return code
+
+def format_with_tabs(code):
+ start = 0
+
+ curly_count, parenthese_count = 0, 0
+ quote_single, quote_double = False, False
+ triple_single, triple_double = False, False
+
+
+ while(start<len(code)):
+ checked_triple = False
+ if(start+3<=len(code)):
+ if(code[start:start+3]=="'''" and not quote_single and not quote_double and not triple_single and not triple_double):
+ triple_single = True
+ start+=3
+ checked_triple = True
+ elif(code[start:start+3]=="'''" and not quote_single and not quote_double and triple_single and not triple_double):
+ triple_single = False
+ start+=3
+ checked_triple = True
+
+ if(code[start:start+3]=='"""' and not quote_single and not quote_double and not triple_single and not triple_double):
+ triple_double = True
+ start+=3
+ checked_triple = True
+ elif(code[start:start+3]=='"""' and not quote_single and not quote_double and not triple_single and triple_double):
+ triple_double = False
+ start+=3
+ checked_triple = True
+
+ if(not checked_triple):
+ if(code[start]=="{" and not quote_single and not quote_double and not triple_single and not triple_double):
+ curly_count+=1
+ elif(code[start]=="}" and not quote_single and not quote_double and not triple_single and not triple_double):
+ curly_count-=1
+
+ if(code[start]=="(" and not quote_single and not quote_double and not triple_single and not triple_double):
+ parenthese_count+=1
+ elif(code[start]==")" and not quote_single and not quote_double and not triple_single and not triple_double):
+ parenthese_count-=1
+
+ if(code[start]=="'" and not quote_single and not quote_double and not triple_single and not triple_double):
+ if(code[start-1]!="\\" or (code[start-1]=="\\" and code[start-2]=="\\")):
+ quote_single=True
+ elif(code[start]=="'" and quote_single and not quote_double and not triple_single and not triple_double):
+ if(code[start-1]!="\\" or (code[start-1]=="\\" and code[start-2]=="\\")):
+ quote_single=False
+
+ if(code[start]=='"' and not quote_single and not quote_double and not triple_single and not triple_double):
+ if(code[start-1]!="\\" or (code[start-1]=="\\" and code[start-2]=="\\")):
+ quote_double=True
+ elif(code[start]=='"' and not quote_single and quote_double and not triple_single and not triple_double):
+ if(code[start-1]!="\\" or (code[start-1]=="\\" and code[start-2]=="\\")):
+ quote_double=False
+ if(parenthese_count==0 and [quote_single, quote_double, triple_single, triple_double]==[False, False, False, False]):
+ if(curly_count>0 and code[start]=="\n"):
+ code = code[:start+1]+"\t"*curly_count+code[start+1:]
+ start+=1
+ start+=1
+
+ return code
+
diff --git a/src/subworkflow.py b/src/subworkflow.py
index b0aed5d57e6e72d7f735f1f2beb823141f39b4c1..1c1ef2c72f85c0fe5b938477db161e597acf2a2d 100644
--- a/src/subworkflow.py
+++ b/src/subworkflow.py
@@ -20,6 +20,7 @@ class Subworkflow(Main_DSL2):
self.call = []
self.initialised = False
+
def set_call(self, call):
self.call.append(call)
diff --git a/src/workflow.py b/src/workflow.py
index 0a0013f1b9d293c5fb3efb826b8e44190c9d3984..9d0f6639b63ee76647abfc416abf6cf0863a7066 100644
--- a/src/workflow.py
+++ b/src/workflow.py
@@ -3,7 +3,7 @@
from .nextflow_file import Nextflow_File
from .ro_crate import RO_Crate
from . import constant
-from .outils import is_git_directory
+from .outils import is_git_directory, format_with_tabs
from .outils_graph import flatten_dico, initia_link_dico_rec, get_number_cycles
from .outils_annotate import get_tools_commands_from_user_for_process
from .bioflowinsighterror import BioFlowInsightError
@@ -82,6 +82,7 @@ class Workflow:
self.dico = {}
self.get_dico()
+
def get_duplicate_status(self):
return self.duplicate
@@ -98,7 +99,9 @@ class Workflow:
self.DSL = DSL
def get_all_executors(self):
- return list(set(self.nextflow_file.get_all_executors()))
+ dico = {}
+ self.nextflow_file.get_all_executors(dico)
+ return list(dico.keys())
def get_is_a_git_repo(self):
return is_git_directory(path = self.get_repo_adress())
@@ -741,43 +744,69 @@ George Marchment, Bryan Brancotte, Marie Schmit, Frédéric Lemoine, Sarah Cohen
def simplify_workflow_code(self):
code, ankers = self.write_workflow_into_one_file()
for exe in self.get_all_executors():
- #print(exe.get_code(), exe.get_type())
- if(exe.get_type()=="Call"):
- print(exe.simplify_code())
- print()
+ if(exe.get_type()=="Call" or exe.get_type()=="Operation"):
+ code = code.replace(exe.get_code(get_OG = True), exe.simplify_code())
+ else:
+ print(exe.get_code(), exe.get_type())
+ raise Exception("This shouldn't happen")
+ return code
#Conert workflow to user_view only makes sense when the option duplicate is activated -> otherwise is doesn't make sense + it makes the analysis way more complicated
def convert_workflow_2_user_view(self, relevant_processes = []):
if(self.duplicate):
+
+ #Write new code in temporary file
+ code = self.simplify_workflow_code()
+ temp_file = self.get_output_dir()/f"temp_{str(self)[-7:-2]}.nf"
+ with open(temp_file, "w") as file:
+ file.write(code)
+
+ #Replace old analysis with new analysis (simplified code)
+ self.__init__(str(temp_file), display_info = False, duplicate=True)
+ self.initialise(create_rocrate=False)
+
+ #for e in self.get_all_executors():
+ # print(e.get_code(get_OG = True))
+
+ #Get the clusters and the code
self.nextflow_file.generate_user_view(relevant_processes = relevant_processes, processes_2_remove = [])
clusters = self.nextflow_file.graph.get_clusters_from_user_view()
print(clusters)
- code, ankers = self.write_workflow_into_one_file()
- #print(code)
-
- for i in range(len(clusters)):
- c = clusters[i]
- if(len(c)>1):
- clusters[i] = self.nextflow_file.graph.get_induced_subgraph(c)
- print(clusters)
+
+ #Get the clsuters with the corresponding operations inside
+ #for i in range(len(clusters)):
+ # c = clusters[i]
+ # if(len(c)>1):
+ # clusters[i] = self.nextflow_file.graph.get_induced_subgraph(c)
+ #print(clusters)
+ #Get the topological order
clusters = self.nextflow_file.graph.get_topogical_order(clusters)
print(clusters)
+
+
+
#Creating the subworkflows from clusters
calls_in_operations = []
for elements in clusters:
+
name, body, take, emit = "", "", "", ""
for ele in elements:
if(ele.get_type()=="Process"):
+
#Determine the name of the created subworkflow cluster
if(ele.get_name() in relevant_processes):
name = f"cluster_{ele.get_name()}"
#Get the call of thing (either process or subworkflow)
#TODO -> check it works with subworkflows
call = ele.get_call()
+
#This verification is really important
if(len(call)!=1):
+ for c in call:
+ print(c.get_code(get_OG=True))
+
raise Exception("This shoudn't happen since duplicate mode is activated")
call = call[0]
printed_condition = " && ".join(call.get_condition().get_conditions())
@@ -786,23 +815,26 @@ George Marchment, Bryan Brancotte, Marie Schmit, Frédéric Lemoine, Sarah Cohen
else:
body+=f"\n{call.get_code()}\n"
elif(ele.get_type()=="Operation"):
- origins = ele.get_origins()
- for o in origins:
- if(o.get_type()=="Call"):
- calls_in_operations.append(o)
- printed_condition = " && ".join(ele.get_condition().get_conditions())
- if(printed_condition!=""):
- body+=f"if({printed_condition}) {{\n{ele.get_code()}\n}}\n"
- else:
- body+=f"\n{ele.get_code()}\n"
+
+ #Ignore these cases
+ if(ele.get_code()[:4] not in ["emit", "take"]):
+ origins = ele.get_origins()
+ for o in origins:
+ if(o.get_type()=="Call"):
+ calls_in_operations.append(o)
+ printed_condition = " && ".join(ele.get_condition().get_conditions())
+ if(printed_condition!=""):
+ body+=f"if({printed_condition}) {{\n{ele.get_code()}\n}}\n"
+ else:
+ body+=f"\n{ele.get_code()}\n"
+ #Here we removing the Call_12313 thing
for call in calls_in_operations:
body = body.replace(call.get_code(), "")
body = body.replace(str(call), call.get_code())
-
subworkflow_code = f"subworkflow {name} {{\n{take}\n{body}\n{emit}\n}}"
- print(subworkflow_code)
- print()
+ print(format_with_tabs(subworkflow_code))
+ print("-----------")
#So basically when retriving a thing (process or subworkflow)