From 19346fe40ac630985e5d00ca2b8e742d572ecb69 Mon Sep 17 00:00:00 2001
From: George Marchment <georgemarchment@yahoo.fr>
Date: Tue, 11 Mar 2025 13:44:29 +0100
Subject: [PATCH] Made the rewrite determinastic -> debugged -> by removing the
zoom opti + removed the artificial nodes from the calculation of the user
view
---
src/graph.py | 3 +-
src/nextflow_file.py | 27 +++++++-------
src/operation.py | 2 +-
src/outils_graph.py | 86 +++++++++++++++++++++++++++++++++++---------
src/process.py | 2 +-
src/workflow.py | 20 +++++++++--
6 files changed, 106 insertions(+), 34 deletions(-)
diff --git a/src/graph.py b/src/graph.py
index c04159d..11d9872 100644
--- a/src/graph.py
+++ b/src/graph.py
@@ -143,6 +143,7 @@ class Graph():
def get_specification_graph(self, dirc = 'graphs', filename = "specification_graph", render_graphs = True):
generate_graph(self.get_output_dir()/ dirc /filename, self.full_dico, render_graphs = render_graphs)
+ generate_graph(self.get_output_dir()/ dirc /(filename+"_without_artificial_nodes"), remove_artificial_nodes(self.full_dico), render_graphs = render_graphs)
def get_specification_graph_wo_labels(self, filename = "specification_graph_wo_labels", render_graphs = True):
generate_graph(self.get_output_dir()/'graphs'/filename, self.full_dico, label_edge=False, label_node=False, render_graphs = render_graphs)
@@ -310,7 +311,7 @@ class Graph():
#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)
- dico = self.dico_flattened
+ dico = remove_artificial_nodes(self.dico_flattened)
user_view, self.new_nodes_user_view = relev_user_view_builder(dico, relevant_modules=relevant_processes)
diff --git a/src/nextflow_file.py b/src/nextflow_file.py
index c19e022..b7a523f 100644
--- a/src/nextflow_file.py
+++ b/src/nextflow_file.py
@@ -160,13 +160,14 @@ class Nextflow_File(Nextflow_Building_Blocks):
#----------------------
#This method extracts the "main" workflow from the file
def extract_main(self):
- from .main import Main
- #This returns the code without the comments
- code = "\n"+self.get_code()+"\n"
- #Find pattern
- twice = False
- for match in re.finditer(constant.WORKFLOW_HEADER_2, code):
- if(self.first_file):
+ if(self.first_file):
+ from .main import Main
+ #This returns the code without the comments
+ code = "\n"+self.get_code()+"\n"
+ #Find pattern
+ twice = False
+ for match in re.finditer(constant.WORKFLOW_HEADER_2, code):
+
start = match.span(1)[0]
end = extract_curly(code, match.span(1)[1])#This function is defined in the functions file
self.main = Main(code= code[start:end], nextflow_file=self)
@@ -174,9 +175,9 @@ class Nextflow_File(Nextflow_Building_Blocks):
#TODO turn into biofow insight error
raise Exception(f"Found multiple 'main workflows' in {self.get_file_address()}")
twice = True
- else:
- #TODO add num
- BioFlowInsightError("A 'main' workflow was found in the Nextflow file")
+ if(self.main==None):
+ raise BioFlowInsightError("A 'main' workflow was not found in the Nextflow file")
+
#----------------------
#FUNCTIONS
@@ -194,8 +195,10 @@ class Nextflow_File(Nextflow_Building_Blocks):
start = match.span(0)[0]
end = extract_curly(code, match.span(0)[1])#This function is defined in the functions file
#f = Code(code=code[start:end], origin=self)
- f = Function(code = code[start:end], name = match.group(2), origin =self)
- self.functions.append(f)
+ #Fobiden names of functions
+ if(match.group(2) not in ['if']):
+ f = Function(code = code[start:end], name = match.group(2), origin =self)
+ self.functions.append(f)
def get_functions(self):
return self.functions
diff --git a/src/operation.py b/src/operation.py
index 1a88d6e..5e838ab 100644
--- a/src/operation.py
+++ b/src/operation.py
@@ -817,7 +817,7 @@ class Operation(Executor):
fillcolor = ""
- dico['nodes'].append({'id':str(self), 'name':"", "shape":"point", 'xlabel': code, 'fillcolor':fillcolor})
+ dico['nodes'].append({'id':str(self), 'name':"", "shape":"point", 'xlabel': code, 'fillcolor':fillcolor, "artificial": self.get_artificial_status()})
for o in self.origins:
#Case origins is a channel
diff --git a/src/outils_graph.py b/src/outils_graph.py
index 17f24dc..0283822 100644
--- a/src/outils_graph.py
+++ b/src/outils_graph.py
@@ -596,14 +596,14 @@ def nr_path_pred(r, n, dico, R):
#Added a dico so it knows what it's already searched
dico_rSucc = {}
def rSucc(n, dico, R, outputs):
- try:
- tab = dico_rSucc[n]
- except:
- tab = []
- for r in set(R).union(set(outputs)):
- if(nr_path_succ(n, r, dico, R+list(outputs))):
- tab.append(r)
- dico_rSucc[n] = tab
+ #try:
+ # tab = dico_rSucc[n]
+ #except:
+ tab = []
+ for r in set(R).union(set(outputs)):
+ if(nr_path_succ(n, r, dico, R+list(outputs))):
+ tab.append(r)
+ dico_rSucc[n] = tab
return tab
def rSuccM(M, dico, R, outputs):
@@ -615,14 +615,14 @@ def rSuccM(M, dico, R, outputs):
#Added a dico so it knows what it's already searched
dico_rPred = {}
def rPred(n, dico, R, inputs):
- try:
- tab = dico_rPred[n]
- except:
- tab = []
- for r in set(R).union(set(inputs)):
- if(nr_path_pred(r, n, dico, R+list(inputs))):
- tab.append(r)
- dico_rPred[n] = tab
+ #try:
+ # tab = dico_rPred[n]
+ #except:
+ tab = []
+ for r in set(R).union(set(inputs)):
+ if(nr_path_pred(r, n, dico, R+list(inputs))):
+ tab.append(r)
+ dico_rPred[n] = tab
return tab
def rPredM(M, dico, R, inputs):
@@ -1082,4 +1082,58 @@ def check_if_equal(dicoA, dicoB):
#TO do that we rewrite the structure using a commun language (without using the ids) -> then just check if the translated structures are the same
return translate_dico(dicoA) ==translate_dico(dicoB)
+#This function removes the artificial nodes from the dico
+#MAtching the nodes together between the artificial nodes
+def remove_artificial_nodes(param_dico):
+ dico = copy.copy(param_dico)
+ def get_list_artificial_nodes_rec(dico, tab):
+ for n in dico["nodes"]:
+ if(n["artificial"]):
+ tab+=[n["id"]]
+ for sub in dico["subworkflows"]:
+ tab = get_list_artificial_nodes_rec(dico["subworkflows"][sub], tab)
+ return tab
+ def get_list_artificial_nodes(dico):
+ tab = []
+ return get_list_artificial_nodes_rec(dico, tab)
+
+ def remove_node_rec(dico, node_id):
+ to_remove = []
+ for n in dico["nodes"]:
+ if(n["id"]==node_id):
+ to_remove.append(n)
+ for n in to_remove:
+ dico["nodes"].remove(n)
+ for sub in dico["subworkflows"]:
+ remove_node_rec(dico["subworkflows"][sub], node_id)
+ def remove_node(dico, node_id):
+ remove_node_rec(dico, node_id)
+
+ def get_links_and_remove_edges_rec(dico, node_id, links):
+ to_remove = []
+ for e in dico["edges"]:
+ if(e["A"]==node_id):
+ links["sink"].append({"sink": e["B"], "label": e["label"]})
+ to_remove.append(e)
+ if(e["B"]==node_id):
+ links["source"].append({"source": e["A"], "label": e["label"]})
+ to_remove.append(e)
+ for e in to_remove:
+ dico["edges"].remove(e)
+ for sub in dico["subworkflows"]:
+ get_links_and_remove_edges_rec(dico["subworkflows"][sub], node_id, links)
+ def get_links_and_remove_edges(dico, node_id, links):
+ get_links_and_remove_edges_rec(dico, node_id, links)
+
+ for n_id in get_list_artificial_nodes(dico):
+ remove_node(dico, n_id)
+ links = {}
+ links["source"] = []
+ links["sink"] = []
+ get_links_and_remove_edges(dico, n_id, links)
+ for source in links["source"]:
+ for sink in links["sink"]:
+ edge = {"A" : source['source'], "B" : sink['sink'], "label" : source['label']}
+ dico["edges"].append(edge)
+ return dico
diff --git a/src/process.py b/src/process.py
index 85c480e..b4cef1a 100644
--- a/src/process.py
+++ b/src/process.py
@@ -416,7 +416,7 @@ class Process(Nextflow_Building_Blocks):
return self.printed_name
def get_structure(self, dico):
- dico['nodes'].append({'id':str(self), 'name':self.get_name_to_print(), "shape":"ellipse", 'xlabel':"", 'fillcolor':''})
+ dico['nodes'].append({'id':str(self), 'name':self.get_name_to_print(), "shape":"ellipse", 'xlabel':"", 'fillcolor':'', "artificial": False})
def initialise_inputs_outputs(self):
DSL = self.nextflow_file.get_DSL()
diff --git a/src/workflow.py b/src/workflow.py
index be488f2..efaab88 100644
--- a/src/workflow.py
+++ b/src/workflow.py
@@ -765,6 +765,7 @@ George Marchment, Bryan Brancotte, Marie Schmit, Frédéric Lemoine, Sarah Cohen
relevant_processes = self.check_relevant_processes_in_workflow(relevant_processes)
self.generate_user_view(relevant_processes = relevant_processes, processes_2_remove = [], render_graphs=render_graphs)
clusters = self.graph.get_clusters_from_user_view()
+ print(len(clusters))
broken_subworkflows = get_workflows_broken(get_subworkflow_2_executors(), get_clusters_with_calls(clusters))
#While there still are broken workflows -> need to redo the analysis
while(len(broken_subworkflows)>0):
@@ -794,6 +795,14 @@ George Marchment, Bryan Brancotte, Marie Schmit, Frédéric Lemoine, Sarah Cohen
#Get the topological order
executors_in_order = self.get_order_execution_executors()
new_clusters = []
+ #for clust in clusters:
+ # print("*")
+ # for c in clust:
+ # if(c.get_type()=="Process"):
+ # print(c, c.get_code()[:20])
+ # else:
+ # print(c, c.get_code()[:20], c.artificial)
+
for cluster in clusters:
tab = []
for e in executors_in_order:
@@ -802,9 +811,14 @@ George Marchment, Bryan Brancotte, Marie Schmit, Frédéric Lemoine, Sarah Cohen
new_clusters.append(tab)
clusters = new_clusters
#clusters = self.graph.get_topogical_order(clusters)
- #for e in executors_in_order:
- # print(e.get_code()[:20])
- #print(executors_in_order)
+ #print('_________________')
+ #for clust in clusters:
+ # print("*")
+ # for c in clust:
+ # if(c.get_type()=="Process"):
+ # print(c, c.get_code()[:20])
+ # else:
+ # print(c, c.get_code()[:20], c.artificial)
#Creating the subworkflows from clusters
--
GitLab