From 561939379d3e08f42fa72b5bdabeefcfa7cfe169 Mon Sep 17 00:00:00 2001
From: tina-z-jia <145156075+tina-z-jia@users.noreply.github.com>
Date: Fri, 6 Oct 2023 16:23:09 -0700
Subject: [PATCH] Update state.py (#1)

* Update state.py

* Update class name to State

* moving example to another file

* default constructor
---
 sanclone/state.py    | 52 +++++++++++++++++++++++++++++++++++++++++---
 tests/test_sanity.py | 25 +++++++++++++++++++++
 2 files changed, 74 insertions(+), 3 deletions(-)

diff --git a/sanclone/state.py b/sanclone/state.py
index dd7b818..657cedb 100644
--- a/sanclone/state.py
+++ b/sanclone/state.py
@@ -1,5 +1,51 @@
-from pydantic import BaseModel
+# -*- coding: utf-8 -*-
+"""get_seq_annotation.ipynb
 
+Automatically generated by Colaboratory.
+
+Original file is located at
+    https://colab.research.google.com/drive/1dJEOj6Jw3qOHsxcsP-W3Uj7mbKhvWnHi
+"""
+
+from Bio import Entrez
+from Bio import SeqIO
+
+
+class State:
+    def __init__(self):
+        self.vector = None
+        self.linear_insert = None
+        self.clone_seq = None
+
+    def store_vector(self, vector):
+        if not isinstance(vector, SeqIO.SeqRecord):
+                raise ValueError("Input 'vector' must be a SeqRecord object from Biopython's SeqIO.")
+        self.vector = vector
+
+    def store_linear_insert(self, linear_insert):
+        # Ensure the linear_insert is a SeqRecord object from Biopython
+        if not isinstance(linear_insert, SeqIO.SeqRecord):
+            raise ValueError("Input 'linear_insert' must be a SeqRecord object from Biopython's SeqIO.")
+        self.linear_insert = linear_insert
+
+    def store_clone_seq(self, clone_seq):
+        # Ensure the linear_insert is a SeqRecord object from Biopython
+        self.clone_seq = clone_seq
+
+    def retrieve_vector(self):
+        return self.vector
+
+    def retrieve_linear_insert(self):
+        return self.linear_insert
+
+    def retrieve_clone_seq(self):
+        return self.clone_seq
+    # seq to annotation
+
+
+def download_genbank_file(accession, filename):
+    Entrez.email = "tina.zetong.jia@example.com"  # Always provide your email address when using NCBI's services
+    with Entrez.efetch(db="nucleotide", id=accession, rettype="gb", retmode="text") as handle:
+        with open(filename, 'w') as outfile:
+            outfile.write(handle.read())
 
-class State(BaseModel):
-    pass
diff --git a/tests/test_sanity.py b/tests/test_sanity.py
index 120d2d1..3cf8939 100644
--- a/tests/test_sanity.py
+++ b/tests/test_sanity.py
@@ -11,3 +11,28 @@ def test_echo_tool():
 
     tool = EchoTool(shared_state=State())
     assert tool.run("Hello") == "Hello"
+
+# def test_state_tool():
+#     from sanclone import State
+#     from sanclone.State import download_genbank_file
+#     accession_id_vector = "NC_005213"
+#     output_filename_vector = "NC_005213.gbk"
+#     accession_id_linear_insert = "NC_000932"
+#     output_filename_linear_insert = "NC_000932.gbk"
+#     download_genbank_file(accession_id_vector, output_filename_vector)
+#     download_genbank_file(accession_id_linear_insert, output_filename_linear_insert)
+
+#     for gb_record in SeqIO.parse(open(output_filename_linear_insert,"r"), "genbank") :
+#         # now do something with the record
+#         print ("Name %s, %i features" % (gb_record.name, len(gb_record.features)))
+
+#     vector_seq = list(SeqIO.parse(open(output_filename_vector,"r"), "genbank"))
+#     insert_seq = list(SeqIO.parse(open(output_filename_linear_insert,"r"), "genbank"))
+
+#     seq_anno = State(vector_seq[0])
+#     seq_anno.store_linear_insert(insert_seq[0])
+
+#     retrieved_vector = seq_anno.retrieve_vector()
+#     retrieved_insert = seq_anno.retrieve_linear_insert()
+#     print(retrieved_vector)
+#     print(retrieved_insert)
\ No newline at end of file